CHILDSPACING OF CURRENTLY MARRIED FILIPINO WOMEN MARRIED ... · CHILDSPACING OF CURRENTLY MARRIED FILIPINO WOMEN MARRIED ONLY ONCE by Jesusa T. Gascon A thesis submitted in partial

CHILDSPACING OF CURRENTLY MARRIED FILIPINO WOMENMARRIED ONLY ONCE

byJesusa T. Gascon

A thesis submitted in partial fulfillment of the requirements for the Degree of Master of Arts in Demography at the

Australian National University Canberra

December 1988

DECLARATION

Except where otherwise indicated, this thesis is my own work

December 1988 sa T. Gascon

iii

ACKNOWLEDGEMENTS

I am most grateful to my supervisor Dr. Alan Gray and my adviser Dr. David Lucas for their guidance, and valuable and constructive criticisms which helped me carry on to the completion of this thesis, and also for sharing with me some reference materials.

I am deeply grateful to the National Statistics Office, on behalf of the Philippine Government and the Australian Development Assistance Bureau for providing me the Colombo Plan Scholarship which enabled me to study at the Australian National University.

I am indebted to the University of the Philippines Population Institute for allowing me to use the 1983 National Demographic Survey and to Mrs. Josefina V. Cabigon for her assistance in creating the computer subfile.

I am sincerely grateful to the MA(D) programme's staff of the National Centre for Development Studies and the staff of the Department of Demography. In particular, I wish to give my deep gratitude to Mrs. Christine McMurray for the assistance she extended in using the mainframe and micro-computers, Mrs. Marian May for correcting my final draft, and to Dr. Christobel Young, I owe my initial idea of using the SPSSX-SURVIVAl software computer package.

Special thanks is also due to Ms. Elizabeth M. Go for her valuable comments and suggestions on the completion of this thesis.

Sincere gratitude also goes to my colleagues and friends particularly Ro-Ann Bacal, Merlyne Paunlagui, Richard Makalew, Mayumi Quintos, Myrna Austria and Helen Cabalu for their friendship and constant encouragement to successfully finish my studies.

Last but not least, to my husband Federico Blancia III, who inspired me and gave me boundless understanding and moral support during the tough years and to my parents, brothers and sister, this thesis I sincerely dedicated.

iv

ABSTRACT

Using 1983 National Demographic Survey (NDS) data on currently married women and been married only once, this study examines childspacing differentials among Filipino women utilizing life table techniques. In particular, the subsequent birth from parity 1 to parity 5 are examined in relation to women's current place of residence, education, contraceptive use, experience of infant mortality and current age.

Examination shows that in Metro Manila and other urban areas, the proportion of women having subsequent birth tended to decline faster after the second birth. In rural areas, the proportion of women having subsequent birth from parity 1 to parity 5 shows only a slight tendency to decline. Compared with rural women, women in Metro Manila and other urban areas have longer birth intervals after the second birth. Younger women have second births sooner but a smaller proportion have subsequent births and they have longer birth intervals at third and higher births than older women. This manifests changing attitudes towards reproductive behaviour among Filipino women.

Results of the analysis confirm the negative effect of education on fertility. Women of higher education exhibit a lower proportion having subsequent births and longer birth intervals after the second birth than women of lesser education. The effect of women's experience of contraceptive use on the quantum and timing of fertility is observed to operate more at the higher parities. In other urban and rural areas, the proportion of ever-users of contraception having further births is lower than for never-users after the second birth but after the third birth in Metro Manila. The study gives further empirical evidence of the effect of infant mortality on fertility behaviour. For each birth order the death of the previous child within the first year of life tended to increase the proportion of women having a subsequent birth, and shortened the average intervals between births.

CONTENTS

v

PageACKNOWLEDGEMENTS ................................. iiiABSTRACT .......................................... ivCONTENTS .......................................... V

LIST OF TABLES .................................... viiLIST OF FIGURES .................................... X

CHAPTER 1. INTRODUCTION .......................... 11.1 Importance and Limitation of the Study .... 21.2 Objective and Organization of the Study ... 51.3 Country Background ....................... 6

1.3.1 Geographic Features ............... 61.3.2 Demographic and Socio-economic

Characteristics ................... 71.3.3 Population Policy and Family

Planning .......................... 131.4 Data Used ................................. 15

CHAPTER 2. REVIEW OF RELATED LITERATURE ......... 202.1 Childspacing Pattern ..................... 21

2.1.1 Developed Countries ............... 212.1.2 Developing Countries .............. 22

2.2 Differentials in Childspacing ............ 252.3 Fertility Differentials Between

Urban and Rural Areas .................... 312.3.1 Education and Fertility ........... 3 32.3.2 Contraception and Fertility ....... 362.3.3 Infant Mortality and Fertility .... 39

vi

PageCHAPTER 3. CHILDSPACING PATTERN AND

DIFFERENTIALS ......................... 423.1 The Methodology Used ...................... 423.2 Childspacing Differences by Place of

Residences ................................ 543.3 Educational Level and Childspacing

Differentials ............................ 583.4 Contraceptive Use and Childspacing

Differentials ............................. 643.5 Infant Mortality and Childspacing

Differentials ............................. 70

CHAPTER 4. EFFECT OF COHORT DIFFERENCES ONCHILDSPACING .......................... 75

4.1 Cohort Differences in Childspacing ........ 764.2 Educational Effect on Childspacing ........ 804.3 Contraceptive Effect on Childspacing ...... 834.4 Infant Mortality Effect on Childspacing ... 87

CHAPTER 5. SUMMARY AND CONCLUSION ................ 915.1 Summary of Findings ..................... 925.2 Concluding Remarks ..................... 98

REFERENCES ......................................... 102APPENDIX - A ........................................ 108APPENDIX - B ........................................ 109APPENDIX - C ........................................ 112

vii

LIST OF TABLES

Table Title Page

1.1 Number of Women in the Sample by Age Groupand Place of Residence ........................ 17

1.2 Demographic Characteristics of the Sampled Currently Married Women, Married Only Onceby Place of Residence, Philippines, 1983........ 19

1.3 Distribution of Currently Married Women,Married Only Once by Parities and Place of Residence, Philippines, 1983.................... 19

2.1 Age-specific and Total Fertility Rates inthe Philippines: 1960-1980...................... 32

2.2 Average Number of Children per Women Aged45-64 by Level of Education and Residence, Philippines, 1973............................... 34

2.3 Percentage Distribution of Currently Married Fecund women According to Use/Non-use of Contraception by Place of Residence,Philippines, 1978............................... 37

2.4 Percentage Distribution of Ever-Married Women Who Have Ever-used Any Contraceptive Method, by Current Age and Family Size,Philippines, 1978............................... 38

2.5 Mean Number of Birth Subsequent to Parity i by Experience of Infant Mortality at thatParity and Residence, Philippines, 1978......... 41

3.1 Percentage Distribution of Currently Married Women, Married Only Once by Age at the Time of the Survey and Place of Residence,Philippines, 1983............................... 46

3.2 Percentage Distribution of Currently Married Women, Married Only Once by Contraceptive Use and Place of Residence, Philippines, 1983. 48

viii

Page

3.3 Percentage Distribution of Currently MarriedWomen, Married Only Once Who Had Experienced Infant Deaths by Age Group and Place of Residence, Philippines, 1983. 50

3.4 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Place of Residence,Philippines, 1983............................... 55

3.5 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Educational Level,Philippines, 1983............................... 59

3.6 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Educational Level and Placeof Residence, Philippines, 1983. 63

3.7 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Contraceptive Use,Philippines, 1983............................... 65

3.8 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Contraceptive Use and Placeof Residence, Philippines, 1983. 69

3.9 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Survivorship of Previous Live Birth in the First Year of Life,Philippines, 1983............................... 71

3.10 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Survivorship of Previous Live Birth in the First Year of Life andPlace of Residence, Philippines, 1983........... 74

4.1 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Age at the Time of theSurvey, Philippines, 1983....................... 77

4.2 Summary Measures for Birth Intervals of BirthOrder 2,3,4,5,6 by Age at the Time of the Survey Controlling by Age at Giving Birth, Philippines, 1983........................ 79

ix

Page

4.3 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Educational Level and Age at the Time of the Survey,Philippines, 1983............................... 82

4.4 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Contraceptive Use and Age at the Time of the Survey,Philippines, 1983............................... 85

4.5 Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Survivorship of Previous Live Birth in the First Year of Life and Age at the Time of the Survey,Philippines, 1983. ........................... 90

Appendix Table1. The Cumulative Proportion of Women of Birth

Order 2,3,4,5,6 who have Subsequent Birth by Birth Interval, Educational Level and Place of Residence, Philippines, 1983................... 109

2. The Cumulative Proportion of Women of Birth Order 2,3,4,5,6 who have Subsequent Birth by Birth Interval, Contraceptive Use and Place of Residence, Philippines, 1983................... 110

3. The Cumulative Proportion of Women of Birth Order 2,3,4,5,6 who have Subsequent Birth by Birth Interval, Survivorship of previous Live Birth in the First Year of Life and Place of Residence, Philippines, 1983................ Ill

X

LIST OF FIGURES

Figure Title Page

1.1 Map of the Philippines Showing PopulationDensity by Region: 1980. ...................... 8

1.2 Age-sex Composition of the Population,Urban-rural, Philippines: 1980. 10

Appendix Figure

1. Life Tables Birth Functions by Parities andPlace of Residence .......................... 112

2. Life Tables by Birth Order, Educational Leveland Place of Residence ...................... 113

3. Life Tables by Birth Order, Contraceptive Useand Place of Residence ...................... 114

4. Life Tables by Birth Order, Survivorship of Previous Live Birth and Place of Residence ... 115

5. Life Tables by Birth Order, Educational Leveland Age at the Time of the Survey ........... 116

6. Life Tables by Birth Order, Contraceptive Useand Age at the Time of the Survey ........... 117

7. Life Tables by Birth Order, Survivorship of Previous Live Birth and Age at the Time ofthe Survey .................................. 118

1

CHAPTER 1 INTRODUCTION

Fertility can be viewed as a sequence of events rather than a single outcome. It is the result of a series of stages in the family building process which starts with the first birth and continues to the end of childbearing. Especially in the developing countries, previous fertility studies often concentrate on the study of various determinants of fertility and fertility levels measured by the number of children ever born or children still living. Insufficient attention is sometimes given to the study of family formation or the process of timing and spacing of births, despite the fact that this sort of study gives a more comprehensive understanding of fertility transition. In fact, the World Fertility Survey series of surveys in the 1970s, among other aims, sought to redress this analytical imbalance.

Previous studies of the Philippines' fertility trend indicate that the country has experienced a declining fertility since the early 1960s. The decline was more visible in the 1970s with the decline of average completed fertility of 6.89 in 1960 to 4.5 in 1978 (Engracia,Raymundo and Casterline,1984:29).

Urban-rural fertility differentials are an established feature in the country. Women in urban areas are observed to exhibit lower and continuously declining fertility while women in rural areas experience high fertility. Factors like education, the use of contraception and infant/child mortality may have contributed to these differentials.

2

Urban-rural fertility studies also indicate the existence of urban-rural childspacing differentials but little is known about this aspect. To demonstrate and further investigate the country's the urban-rural childspacing pattern and effect of cohort difference on fertility behaviour are the aims of this research.

1.1 Importance and Limitation of the StudyIn the studies of childspacing there are two

interrelated components of fertility processes involved. One aspect refers to the quantity (quantum or intensity) of fertility, which can be measured by the proportion of women at each parity who eventually move to the next highest parity (the parity progression ratio). The other aspect is the timing (or tempo) of fertility, which can be measured by the time it takes to make the transition from one parity to the next for those women who continue reproduction, or the distribution of birth intervals (Rodriguez and Hobcraft,1980:8). Changes in either or both of these quantum and tempo measures of fertility alter the fertility level and eventually affect the fertility pattern. A study on changes in childspacing practices is important for it explains the pace of fertility in society over time and considering differentials in socio-economic and demographic characteristics can pave the way to understand this fertility behaviour.

A parity progression ratio can delineate childspacing practices by describing the proportion of women who have already reached parity n and then further proceed to parity n + 1, and if calculated from different cohorts of women it

3

gives relevant measures of fertility behaviour. Estimates of parity progression ratios obtained from older women indicate childbearing experience up to 20 years earlier while estimates derived from younger women who are still bearing children gives the recent fertility behaviour. Long term fertility trends can also be depicted by looking at the proportion of women who have completed childbearing by parities.

Lack of suitable data could be one of the reasons for the existence of few studies on spacing patterns and changes, particularly in developing countries. There have, however been substantial progress on the analysis of socioeconomic determinants of birth interval dynamics in the past 10-15 years. In large part, this advancement on birth interval analysis arise out from the results of the World Fertility Survey (WFS) participated by different countries. Fertility behaviour studies emanated from the WFS data, for examples are : Rodriguez and Hobcraft's (1980) birth interval analysis of Columbia; Feeney's (1983) studies of population dynamics based on birth intervals and parity progression; Rindfuss et al's (1984) analysis of childspacing in Asia; and Guilkey's (1988) study on the effect of current characteristics and rural development on childspacing in the Philippines.

Longitudinal surveys that give complete birth history information are seldom conducted because they are time consuming and more expensive unlike cross-sectional surveys. Though cross-sectional surveys give good birth history information through improved survey techniques, data from such fertility survey obtains incomplete birth

4

histories for younger cohorts who have not yet completed childbearing. This leads to selectivity and censoring biases. Selectivity bias occurs when women tend to be selected on a number of characteristics and are not representative of the total population. For instance, the fact that the transition from parity n to n+1 can only be studied for women who have reached parity n or more at the survey date introduced selectivity bias. Also, the study only included women who were still alive and still married. Censoring bias, on the other hand, occurs when not all women have completed their childbearing and exposure to having the next birth is truncated by the date of interview. This tends to underestimate the length of the birth interval since some of the open intervals may eventually be closed some time in the future

Nevertheless, as stated by Srinivasan (1980), censoring bias can be handled using life-table techniques. On the other hand selectivity bias can be tackled, though not completely, by constructing separate life tables for women reaching each parity at various ages at time of interview, as suggested by Rodriguez and Hobcraft (1980). The justification is that although the life-table technique cannot estimate the parity progression ratio, it does estimate the proportion of all women in the sample who have a subsequent birth at successive durations of exposure.

5

1.2 Objectives and Organization of the StudyTo examine differentials in childspacing pattern in

urban and rural areas in relation to fertility changes is the primary objective of this study. The study will look at the overall picture of the pace of fertility change among Filipino women living in the most urbanized area (Metro Manila), other urban areas and rural areas. It will further examine the effect of cohort difference on fertility behaviour, and difference in fertility by educational level, contraceptive use and experience of infant mortality.

This study is made up of five chapters. The remainder of this chapter discusses the geographic features of the Philippines, demographic and socio-economic characteristics, and the country's population policy. This chapter also give brief description of the survey which is the source of this study's data and the sample's characteristics. Chapter 2 discusses a review of related literature on patterns and differentials in childspacing in both the developed and developing countries. The fertility differentials of Filipino women according to differences in their place of residence, educational level, contraceptive use and experience of infant mortality are also discussed. Chapter 3 has five sections. The first section give details of the study's methodology, its dependent variables, its independent variables and measures used in the study. The succeeding sections examine the childspacing differentials among women of different socio-economic characteristics. Chapter 4 examines the cohort effect on childspacing differentials between Filipino women. Effects of

6

educational level, contraceptive use and infant mortality- are examined again with age being controlled. Lastly, Chapter 5 give a summary and conclusion.

1.3 Country BackgroundThe principal sources of this subsection are the

Statistical Handbook of the Philippines 1984 (NCSO,1984), National Summary of the Philippines 1980 Census of Population and Housing (NCSO, 1983), An Introduction Guide to Population and Development Planning (NEDA,1983), and the RPFS-WFS First Report (NCSO et al.,1979).

1.3.1 Geographic FeaturesThe Philippines, with 7,100 islands, is bounded in the

west by the South China Sea, in the east by the Pacific Ocean, in the south by the Sulu and Celebes Seas, and in the north by the Bashi Channel. It has a total land area of 300,000 square kilometres which is divided into three major island groups: Luzon, Visayas and Mindanao. The biggest island, Luzon, accounts for 143,395 square kilometres and Mindanao, the second largest, for 101,999 square kilometres. With lofty highlands and numerous valleys, the country's interior is generally mountainous. These mountain ranges traverse the country forming natural barriers. Rivers and streams which run through the country are used for irrigation and transportation.

Administratively, the three island groups are further subdivided into regions, regions into provinces, provinces into cities, municipalities, and municipal districts. Cities, municipalities and municipal districts are divided

7

further into barangavs. the smallest administrative divisions of the country. As of March 1980, there were 13 regions including Metropolitan Manila, 73 provinces, 60 cities, 1,484 municipalities, 21 municipal districts and 39,926 barangavs. Metropolitan Manila is also known as the National Capital Region (NCR) made up of 4 cities (Manila, Caloocan, Pasay and Quezon City) and thirteen municipalities (see Figure 1.1).

The City of Manila, the official capital of the Philippines, is the country's main port of entry, the hub of trade and commerce, and the seat of cultural and intellectual activities. Being a gateway to the various islands and the center of Philippine life, Manila is the heartland of the entire archipelago.

1.3.2 Demographic and Socio-economic CharacteristicsThe total population as of May 1, 1980 was

48,098,460 averaging an annual growth rate of 2.71 percent from 1975. The population density of 140 persons per square kilometre in 1975 increased to 160 persons per square kilometre in 1980. The most densely populated area, Metro Manila, had a population density of 9,317 persons per square kilometre. Also, as the country moves towards industrialization, the concentration of urban population increases. From an urban share of 31.6 per cent of the population in 1975, it increased to 37.2 per cent in 1980.

8

F I GURE 1 . 1MAP OF THE P H I L I P P I N E S SHOWING

POPULATI ON D EN S I T Y BY REGION: 1 9 8 0

POPULATION DENSITY PER SQUARE KM.

NATIONAL CAPITAL REGION METROPOLITAN MANILA9,OOC ond over

Under 7 4

■ LOCOS REGION

2 CAGAYAN VALLEY

CENTRAL LUZON

5 3ICOL REGION

A v /

SOUTHERN TAGALOG 4

EASTERN VISAYAS

WESTERN VjISAYAS 6 K 'c CENTRAL VISAYAS

n o r t h e r n

MINDANAO

WESTERN MINDANAO 9 SOUTHERNMINDANAO

d x ? CENTRAL MINOANAO

S o u r c e s : NCSO, 1984: F i g u r e 3 .1 and NCSO, 1983.

9

The Philippines population has an extremely young age composition and a small proportion of old people as depicted by the population pyramid in Figure 1.2 which has a broad base and low gradient sides indicating a high rate of fertility which is fairly constant. The population under 15 years of age in the 1980 Census of Population constituted 42.0 per cent of the total while those aged 65 years and over was 3.4 per cent. This made the dependency ratio 8 3.3 compared to 93.9 twenty years ago. The median ages of 17.6 also confirms that the country has a relatively young population but when compared with the 1970 and 1975 median age of 16.4 and 15.8 respectively, it may be concluded that it is progressively ageing.

Males slightly outnumbered females with the sex ratio for the country as a whole 100.7 in 1980 compared to 102.3 in 1975 and 99 in 1970. However, females outnumbered males in these three regions: Metro Manila, Ilocos and Central Visayas. Metro Manila registered the lowest sex ratio - 93 males for every 100 females. Generally, more females than males lived in urban areas while in rural areas the opposite was the case.

Marriage has remained a typical phenomenon in the Philippines setting. Although the proportion of women who never marry has remained practically the same over the last seven decades, the average age at marriage significantly increased by 2.5 years from 1903 to 1973 (NCSO et al., 1979) . In terms of singulate mean age at marriage (SMAM), the change in the marital distribution between 1903 to 1975 indicates a postponement of marriage by about 2.3 years.

10

FIGURE 1.2 AGE-SEX COMPOSITION OF THE URBAN-RURALPOPULATION, PHILIPPINES:1980

AGE GROUP ( IN YEARS) 80 AND OVER

7 3 - 79

70 - 7 4

«5 - 69

80 - 64

58 - 39

30 - 84

4 5 - 4 9

4 0 - 4 4

3 5 - 39

30 - 34

2 5 - 29

2 0 - 24

I 6 - 19

10 - 14

5 - 9

UNDER 5 I

9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8

PERCENT TO TOTAL POPULATION

Source: NCSO, 1983: Figure 1.

The population growth of the Philippines is mainly attributed to the excess of births over deaths, for migration is comparatively insignificant. The crude birth rate, which was fairly constant during the first half of the century, ranged from 50 to about 56 per thousand of the population (NEDA, 1983: Table 11.6:13). A declining trend in the crude birth rate, nevertheless, has been apparent since the 1950s (NEDA, 1983:13). During the mid-1950s the estimated crude birth rate was placed around 49 to 52 per thousand population, suggesting a decline from previous levels of about 6 to 7 per cent. Further decline has been observed in the 1960s with the birth rate estimated at around 4 2 to 4 6 per thousand population. For the period 1970-1975 the birth rate is estimated on average at 37.4

11

(UN, 1978:117). The fertility estimates for the country in 1980 place the crude birth rate at 33.2 per thousand population (NEDA, 1983:14). This decline is believed to be the result of delayed marriage offsetting the slight increase in marital fertility because of improved health and a shortening of inter-birth intervals (UN,1978:118).

Declining mortality has been observed in the country since the Second World War. The estimated death rate of about 20 per thousand in 1950 dropped to only 13 in 1960 (NEDA, 1983:12). Although at a moderate pace, further improvements continued in the years that followed placing the crude death rate at 8.3 in 1980. Overall improvement in mortality can be said to be the result of the reduction in infant mortality. The 1960 infant mortality estimate of 1 out of 10 live births has been reduced to somewhere about 60 to 65 deaths per 1000 live births. As with fertility, mortality variations exist between regions and among different subgroups of the population. Notably, rural areas exhibit higher fertility and mortality rates than urban.

A salient feature of the Philippine culture is the diversity of languages. Of more than 87 languages anddialects in the country enumerated during the 1980 Census of Population and Housing, ten are the mother tongue of over 89.5 per cent of the population of 1980. Among the first three languages spoken are Tagalog, Cebuano and Ilocano. A disseminated form of the Tagalog language called Pilipino is the national language, and English is the other official language in the country. The average literacy rate among the population 10 years old and over

12

was quite high at 82.7 per cent in 1980 with the male literacy rate (82.8 per cent) not greatly different from that of females (82.6 per cent). However a marked difference in literacy rate was observed between urban’ and rural residents, 92.5 per cent for the former and 7 6.5 per cent for the latter. In the 1980 Census, 9.5 per cent of the population 7 years old and over had no grade completed, 56.3 per cent had elementary, and 3 3.8 per cent had high school or higher education. There are about 0.3 per cent "not stated" cases.

Having been under different colonial powers, the Philippines adopted several religions. Most predominant is Christianity which in itself is composed of various sects, although most Filipino Christians are Roman Catholic. Protestants include Baptists, Evangelicals, Lutherans and a host of others. Local variants are the Iglesia ni Kristo and the Aglipayans (Philippine Independent Church) while a small proportion of the population embraces Buddhism. The 1970 census on religion disclosed that 85 per cent of the country's population are Roman Catholics, 4.3 pe cent are Muslims, 3.9 per cent are Aglipayans, 3.1 per cent are other Protestant, 1.3 per cent are Iglesia ni Kristo and 0.1 per cent are Buddhist.

Extended family structure and close family ties describe the Filipino society. Though the degree of family coherence differs between urban and rural families and between the rich and the poor, it plays a vital role in various aspects of social, economic or even political activities. In general, children are expected to look after their parents in the future.

13

Filipino women are considered to play an important part in economic and political decision-making. Compared with other Asian countries, they have relatively high labour force participation. Of the population aged 15 years and over in 1980, the male labour force participation rate was 79.1 per cent while that of the females was 45.7 per cent. Of the total work force of 18.3 millions in 1980, 47.5 per cent were agricultural, animal husbandry and forestry workers. Although noticeable growth has occurred in the Philippines economy over the past years, poverty stemming from income inequality still exists.

1.3.3 Population Policy and Family PlanningWith the belief that the greatest welfare can be

achieved through non-interference with family size because childbearing is essential and natural, the Philippines laws with some bearing on population growth favored pro-natalist during the 1960s (Peralta and Ligan, 1975:2). It was not until 1969 that the Philippines government made an effort to control population trends for national and social planning. In the same year the Commission on Population (POPCOM) was created to study, reformulate, adopt and plan a population program related to economic and social developments. The next year, 1970, the opening speech of the country's President to the Seventh Congress made family planning the official policy of the administration (Concepcion,1978:173).

The family planning programme has identified four areas of activity (Concepcion,1978:173):

a) delivery of all family planning services through

14

clinics situated all over the country;b) training of all types and level of personnel in

the programme;c) information, education and communication; andd) research and evaluation.

It is a non-coercive policy, based on the voluntary act of the couple to decide and choose among contraceptive methods.

The Philippines government under its Development Plan 1984-1987, aspired to reduce annual growth rate of 2.54 per cent in 1983-1984 to about 2.37 per cent by the end of 1987. The 1983 crude birth rate and the crude death rate of 33.6 and 8.2 per thousand population, respectively, are expected to be lowered by 1987 to 31.3 and 7.6 per thousand population respectively. On the other hand, thecontraceptive prevalence rate estimated at around 34 in 1983 is expected to rise in 1987 to 40.5 per cent (NEDA,1984:230). However, contrary to expectation, the 1983 National Demographic Survey (NDS) showed that there has been a reduction in the proportion of married women of reproductive age practising family planning. Theprevalence rate of 37 per cent in 1978 has been reduced to 32 percent - a little lower than the estimated prevealence rate from the 1983 NDS (RP, 1987:221) . In effect the marital fertility rate did not decline.

Knowing this situation, the government's Philippine Development Plan 1987-1992 has set out specific policies and strategies to meet its new target. The target is to reduce the total fertility rate from 4.1 in 1987 to 3.7 in 1992. Its policies include: the promotion of a family

15

planning program that concerned not only method of contraception but also maternal and child health; and the enhancement of education for the population, especially the adolescent group, on responsible parenthood (RP, 1987:231).

1.4. Data UsedThis study utilizes the 1983 National Demographic

Survey (NDS), the most recent available national demographic survey in the Philippines. The surveyquestionnaire that was given to ever-married women (EMW) 15-49 years, as the eligible respondents, consisted of nine blocks of information, but only five blocks were used in this study. These were:

A. IdentificationB. Household QuestionnaireC. Pregnancy HistoryE. Contraceptive Knowledge and Use and Fertility

PreferencesG. Nuptiality

The 1983 NDS collected information from a nationally representative sample of 13,000 households. An average of 1,000 households was selected per region to permit estimates at regional level. The sampling design consisted of stratified two-stage sample wherein the primary sampling units, the barangavs. were selected with replacement and with probability proportional to the number of households per barangav. The secondary sampling units, thehouseholds, were sampled systematically with a random start. The sample was self-weighting within regions, that

16

is equal numbers of households were selected within sampled barangavs.

In the survey a total of 10,845 EMW were successfully interviewed. However, this study focuses only on the 8,933 currently married women who were married only once and did not experience multiple births. The reason behind this was to remove certain biases introduced in the computation of the interval between live births by EMW who already lost their partner or women who remarried and resumed giving birth, and for less complication in computation of birth intervals. Furthermore this study excludes the first birth interval, the interval between marriage and first birth. Bogue and Bogue (1980) concluded in their birth interval analysis, that the first birth should be treated separately and differently from other succeeding births. Unlike with other birth intervals where date of prior birth is used as the initial point for the start of the interval , first birth uses date at marriage as the reference point for start of exposure to childbearing. In addition, some factors that affects second to higher birth intervals does not apply to first birth. For examples, factors such as duration of postpartum amenorrhea and breastfeeding. While omitting the first birth interval produces a less complete analysis, it was decided to avoid the analytical complication in this thesis.

Number of women by five years age-group is showed in Table 1.1. It can be seen from the table among the 429 ever-married women of age-group 15-19, only 57.6 per cent (247) of them had first marriage still intact. In the oldest age-group (44-49 of age), however, 3 out of 4 (1053

17

out of 1406) ever-married women are still in their first marriage. The bulk of currently married women still in their first marriage consisted of women in the age-groups 20-24 to age-group 40-44.

Table 1.1 Number of Women in the Sample by Age Group and Place of Residence

Characteristics/ Place of Residence 15-19 20-24

Age of Women 25-29 30-34 35-39 40-44 45-49

Total

All ever-married 429 1633 2172 1959 1624 1622 1406 10845Metro Manila 38 234 294 292 218 218 168 1462Other Urban 84 367 562 505 388 409 346 2661Rural 307 1032 1316 1162 1018 995 892 6722

Currently married* 247 1338 1891 1728 1368 1308 1053 8933Metro Manila 26 181 223 247 178 162 129 1146Other Urban 46 287 476 433 319 325 261 2147Rural 175 870 1192 1048 871 821 663 5640

Currently married with atleast one live birth* 227 1311 1863 1713 1359 1301 1047 8821

Metro Manila 25 181 221 243 178 162 127 1137Other Urban 44 280 468 427 319 322 258 2118Rural 158 850 1174 1043 862 817 662 5566

* included women are married only once

Source: Compute from the 1983 National Demographic Survey (NDS) data tape.

Table 1.2 shows the demographic characteristics of the women in the sample under study. Consistent with other studies, it can be seen that women residing in Metro Manila and other urban areas have higher education than the rural women. It also shows that women living in Metro Manila and other urban areas have less experience of child death, a lower currently pregnant rate but a higher level practicing family planning and experience of miscarriage than those women in the rural areas. Distributions of women by parities and place of residence are shown in Table 1.2. It

18

can be seen from the table that as the parity progresses to higher order, the number of women decreases. This is from the fact that women who attained the parity n may or may not proceed to parity n + 1. This study focuses only on childspacing practices among women of parity 1 to parity 5 and the number of women in each life table may vary depending upon the completeness of information of the specified variables.

Table 1.2 : Demographic Characteristics of the Sampled Currently Harried Women, Married Only Once by Place of Residence, Philippines, 1983

(in per cent)

Characteristics Metro Manila

OtherUrban

Rural Total

With no live birth 0.9 1.4 1.2 1.2

With only one live birth 17.4 15.8 12.0 13.6

Experiencing still birth/miscarriage 23.7 26.0 23.0 23.8

Experiencing death of a child 17.5 20.1 24.6 22.6

Currently pregnant 8.2 9.5 11.3 10.5

Currently practicing familyplanning 47.2 38.4 26.9 32.2

Work statusworking 37.6 32.9 20.4 25.6not working 62.4 67.1 79.6 74.4

Highest grade completednone/primary 6.4 13.6 25.7 20.3intermediate 24.8 31.4 42.3 37.5HS or college 68.9 55.0 31.9 42.2

Number of women (1147) (2147) (5639) (8933)

Source: Computed from the 1983 National Demographic Survey (NDS) data tape.

Table 1.3 : Distribution of Currently Married Women, Married Only Once by Parities and Place of Residence, Philippines, 1983

Number of Eligible WomenParity

Metro Manila

OtherUrban

Rural Total

0 1,147 2,147 5,639 8,9331 1,137 2,118 5,566 8,8212 937 1,779 4,890 7,6063 696 1,411 3,996 6,1034 465 1,022 3,112 4,5995 305 692 2,323 3,320


20

CHAPTER 2REVIEW OF RELATED LITERATURE

Several studies point to a declining fertility level in the Philippines, emerging most probably from changes in the reproductive behaviour of Filipinos in past decades and the adoption by the government of a national population policy and implementation of supporting programs in 1970. The decline of fertility occurred throughout the country, though it took place at different times and speeds in the regions and urban and rural areas. Studies have also shown that fertility is lower in more urbanized areas such as Metro Manila than other less urbanized or rural areas. In this context, the present study examines fertility behaviour in terms of childspacing practices among Filipino women with relation to their place of residence. This chapter presents a review of relevant literature and is divided into two main sections. Patterns and differentials in childspacing in both developed and developing countries are discussed in the first section while the second section focuses on urban-rural differentials in fertility in the Philippines. The second section also discusses differences between urban and rural characteristics in terms of level of education, contraceptive use, and experience of infant mortality and its effect on fertility.

21

2.1 Childspacing Pattern

2.1.1 Developed CountriesRetrospective studies in some developed countries have

found a faster pace of childbearing in more recent marriage cohorts and a shorter duration of the childbearing period. As noted by Young (1977) in her study of Australian women married between the period 1905-1909 and 1965-1969, this tendency can be explained by the decrease in the average family size and a closer spacing of births since the Second World War.

Farid (1974) claimed that the high fertility rates observed during the period 1956-1964 in England and Wales, were the result of the changes in the quantum of fertility as well as the tempo of childbearing. The fertility of marriage cohorts in these regions during that period was characterized by a decline in the frequency of childless marriages and by a dramatic decrease in the proportion of marriage cohorts with two or more children. Moreover, Farid's study concluded that the diffusion of the 2-3 child family among marriage cohorts between the middle 1950s and the early 1960s was accompanied by a strong tendency for childbearing to start earlier in marriage and be concentrated into a shorter period. These factors compressed the effective childbearing period for these cohorts, causing a contraction in the mean length of interval between marriage and successive births.

22

Klinger (1977) found the same pattern of childbearing in his study of marital fertility schedules of Hungarian women. He found that childbearing was taking place early in the reproductive span, and women were having their second child earlier, that is the second birth intervals were becoming shorter.

A similar pattern of childbearing and completed fertility took place in the United States in the 1950s and the early 1980s. Thorton and Freedman (1983) concluded that the substantial drop of total fertility from 1957 to 1982 reflected the American couple's preference for fewer but closely spaced children. In fact Whelpton (1964) had already observed these patterns in his study in 1960, "Growth of American Families" (GAF). The GAF study shows that the mean length of interval between births decreased rapidly from the birth cohorts of 1916 - 1920 to those of 1936 - 1942.

2.1.2 Developing CountriesThough there have been few cohort studies in spacing

in the fertility patterns in Asian countries, that fertility patterns among developing countries are said to be responding to their rapid socio-economic growth and modernization (Merrick et al., 1986: 37). Some Asiancountries are experiencing a decline in fertility with the pattern of reproduction more likely to follow the same pattern observed in the developed countries - the pattern of early first birth and the rapid succession of subsequentbirths.

23

An extreme example is Singapore where a dramatic drop in fertility occurred concurrent with remarkable economic and social development. Fawcett and Khoo (1980) noted the country's changing attitudes, values and aspirations for children favouring a norm of small family size for the 15 years since its independence. According to the same study, the span of childbearing tends to be shorter among Singaporean women in recent years than the 30 years biological span of childbearing. This is because most women tend to enter a marital union at a much later age. [Eventually, this may lead to closely spaced subsequent births for women to make up the loss of some years of childbearing].

Donaldson and Nichols (1978) studied marriage cohorts of Korean women born in the period 1936-1965. They found, among Korean women, a trend towards a more rapid pace of childbearing and a greater concentration of fertility at early durations of marriage. They found that more recent cohorts had increases in marital fertility at younger ages. Further, not only did recent marriage cohorts have their first birth sooner after marriage, they also had subsequent births sooner after the first birth than did women in the older marriage cohorts.

Kandiah (1981) found that there is a concentration of childbearing in the early years of marriage - during the first five years of marriage - among Malaysian women. This is particularly evident, as Chandler et al. (1976) noted, among Malaysian women who marry at later ages. A similar

24

trend of childbearing was found among Taiwanese women (Freedman and Casterline, 1979).

In the Philippines, Cabigon and Hufana (1984), using the 1978 Republic of the Philippines Fertility Survey, concluded that little effort was exerted among Filipino women to delay the onset of childbearing after marrying. In fact, Cabigon (1984) in her study of the timing of first births in the Philippines, concluded the main factor delaying the first birth is marriage. However, the interval between births increases as women progress to higher parity, particularly after the third birth.

Effects of current characteristics and rural development on childspacing in the Philippines was analyzed by Guilkey et al. (1988) . The study used the Bicol Multipurpose Surveys of 1978 and 1983, and utilized the discrete-time form of Cox's proportional hazards model as the method of analysis. The study concluded that family building process can be influenced by both fixed and changing socio-economic characteristics. Further, the study educe effects of changing socio-economic characteristics on family formation could be traced through changing couples' decision about contraception, abortion or sterilization. It also showed the possible effects on birth intervals of changing socio-economic characteristics on the decision in other areas such as breastfeeding, spousal separation or possibly various infectious diseases.

25

2.2 Differentials in ChildspacincrDeterminants of fertility, and birth intervals can be

grouped into two categories - biological and socio-economic determinants. Biological determinants are factors affecting the physiological capability of the women to conceive. Socio-economic determinants are factors affecting fertility indirectly by influencing the set of intermediate variables embodied in the biological factors. The three sets of intermediate variables as classified by Davis and Blake (1956) were:

(a) factors affecting exposure to intercourse,(b) factors affecting exposure to conception, and(c) factors affecting gestation and successful

parturition.The effect of these intermediate variables on fertility can be positive or negative depending on the effects of environmental, social, and economic factors on these variables.

Modernization and development would affect some of the intermediate variables through changes in socio-economic factors which may affect the intermediate variables directly or through their substantial influence on norms of family size and norms about the intermediate variables. In India, for instance, Srininivasan, Reddy and Raju (1978:264) in their study of changes in fertility from one generation to the next found that

...the reduction in birth intervals seems to be traceable to both cultural and technological effects of modernisation. Intervals between successive births in India have in the past been consistently long due to prolonged lactation and low fecundability caused by

26

cultural and religious taboos regarding sexual intercourse and possibly due also to malnutrition and disease.

In terms of the effects of intermediate variables on spacing of births within the fertility span, the two factors related to the process of modernization that are important are duration of breastfeeding and duration of postpartum abstinence. Likewise, some socio-economicdeterminants of birth intervals like education, contraceptive use and infant and child mortality are also related to the process of modernization and development.

Post-partum Amenorrhea and BreastfeedingIn a "traditional" society where modern methods of

birth control are not much in practice or are totally absent, the length of the birth interval is determined largely by the period of postpartum amenorrhea which can be lengthened through breastfeeding. The median postpartum amenorrheic period of two months can be extended to an average of one to two years through prolonged and intensive breastfeeding (Go, 1983: 2-3). Van Ginniken et al.'s(1978) study in rural areas of some developing countries found that lactational amenorrhea increased birth intervals by more than 15 months which accounted for about 4 0 per cent of the average birth interval. In the countries of Latin America and the Carribean, where women breastfeed only for a short period of time, postpartum amenorrhea does not much influence birthspacing; other factors are

27

responsible, for instance, the practice of contraception (Ferry and Smith, 1983: 4).

In Bangladesh, Pakistan, Nepal and most of Sub-Saharan Africa, where very few women practice contraception, prolonged breastfeeding has kept fertility rates from becoming even higher than they are (Merrick et al., 1986: 31). In fact in Bangladesh, Arthur and McNicholl (1978) concluded that breastfeeding is the main reason for the country not reproducing at the level of Hutterite society. They also claimed that lactational amenorrhea accounts for nearly 19 months of the average 34-month birth interval in the country.

Ferry and Smith (1983) cited some factors affecting women's duration of breastfeeding. One factor concerned the mother's parity, which is not independent of the woman's age; it is said that some women breastfeed their first children for a significantly shorter period than their subsequent children. Another factor is the woman's place of residence. They found that mothers inmetropolitan areas tend to breastfeed their child for a shorter period than mothers in other towns or rural areas. Thirdly, factors such as education, which is notindependent of the type of residence and other socioeconomic variables, also play an important part in the reduction of a woman's length of breastfeeding. Lesthaeghe et al.'s (1981) study of women in Lagos, Nigeria, shows that less educated mothers have a longer duration of breastfeeding than those with higher education. Lastly,women's place and nature of work is said to affect

28

breastfeeding significantly. It is most likely that women working outside and far from home find it moreinconvenient to breastfeed.

In addition, the introduction of prepared commercial infant foods enables women to wean their babies earlier. This trend, though convenient for women in terms of time, has an impact on fertility and infant mortality. Knodel (1977) concluded that apart from increasing women's exposure to the risk of pregnancy, substitution of commercial infant foods for breastmilk also increases infant mortality.Postpartum Abstinence

In some societies not only is postpartum amenorrhea affected by breastfeeding but also the practice of postpartum sexual abstinence. Among the Yoruba, in Nigeria postpartum sexual abstinence is traditionally practiced as long as the mother is breastfeeding and thus it can be extended to two or three years (Santow, 1978: 94).Eventually, since the absence of ovulation during postpartum amenorrhea lasts for a shorter time than the lactational period, childspacing is achieved through the practice of post-natal sexual abstinence.

In some societies (among Ibadan women of the western states of Nigeria, for example) the period of abstinence exceeds the period of breastfeeding and has the effect of keeping fertility levels lower than they would be otherwise (Caldwell and Caldwell, 1981; Santow, 1978). The fear that the mother's milk may be contaminated and the need to safeguard the mother's health are among the reasons why

29

sexual abstinence is practiced during the period of lactation in this society. In most African societies, factors to do with morality and public criticism were among the additional reasons for the practice of abstinence (Santow, 1978).

Infant and Child MortalityIt is postulated that infant and child mortality

directly affects fertility in two ways. The first is that it tends to increase the number of children ever born because women tend to replace their children who have died. Another is that it tends to shorten the intervals between births because of reduced periods of breastfeeding and amenorrhea (Rodriguez and Hobcraft, 1984: 34). As a corollary to this, as has been argued under the "child- survival hypothesis", improving child survival would increase family planning motivation and consequently lead to fertility decline (Taylor and Hall, 1967; Taylor et al., 1976). The association between birthspacing and infant mortality is generally strong among countries (Middle East and North Africa, for instance) that mostly depend for the timing of the next birth on natural birthspacing: breastfeeding, amenorrhea, and abstinence (Pebley and Millman, 1986: 74). Infant or early childhood mortality shortens the period of natural birthspacing in these countries.

30

Contraceptive UseA general impression has been built up that

childspacing has always been an important motive in the use of contraception. Hobcraft and McDonald (1984) provided indirect but fairly convincing evidence that contraceptive spacing has lengthened lower order birth intervals in Latin America. However, in Asian countries, perhaps because of the greater contribution of breastfeeding to birthspacing, contraceptive demand appears to be dominated initially by the desire to have no further children (Merrick et al., 1984: 34; Cleland and Rutstein, 1986). A study of birth intervals using World Fertility Survey data, shows little use of contraception for childspacing in Asia compared with Latin America (Cleland and Rutstein, 1986). London et al. (1985), in a study based on 10 national surveys in Asia, show that on average 3 3 per cent of couples use contraception to delay the next birth and 55 per cent use contraception to prevent additional births.

Differences in childspacing practices were also observed among women of different ages at marriage. Generally, women who marry at older ages have higher fertility (or shorter birth intervals) at early duration of marriage. Some of the possible reasons for this trend are a decline in the duration of breastfeeding and the proportion of women breastfeeding, increased coital freguency in the early years of marriage to " catch-up" for the lost years of childbearing, and possibly higher fecundity among higher status groups who are more likely to marry late (Kandiah, 1981).

31

Moreover, childspacing practices were observed to vary among women of different levels of education. Mosley et al. (1982) concluded that, in the short run, an increase in female education can actually raise fertility and shorten the gap between births as more educated women often abandon traditional practices which have fertility-suppressing effects such as breastfeeding and post-partum abstinence. Nevertheless, this effect had been fully compensated for at the same time as more educated women exhibit a higher level of contraceptive practice either to space or limit births.

2.3 Fertility Differentials Between Urban and Rural AreasMost studies trying to unfold fertility differentials

reach the conclusion that fertility is largely affected by socio-economic and cultural factors. In any society, where socio-economic and cultural differences exist, no doubt fertility differentials will be observed. Within a country, some of the most marked socio-economic and cultural differences are those found between urban and rural areas. Hence, differences in fertility behaviour between urban and rural population are expected to emerge.

Fertility studies in the Philippines concluded that the country experienced a decline in fertility since the early seventies (de Guzman, 1978: 118? NEDA, 1983: 13). As Table 2.1 shows, at the peak of childbearing, which is at ages 25-29, the age-specific fertility rate dropped by 22 per cent from 1970 to 1980. In terms of the total

32

fertility rate, the data suggest that in 1960, on the average woman could end up with a completed family size of 6.5 children at the rates observed in that year while in 1980, she could expect to have only 4.5 children. This is a drop of 2 children per woman during the 20-year period. Existed decline occurred in the 1970s when the expected completed family size dropped by about 1.4 children (from 5.89 to 4.48) .

Table 2.1 : Age-Specific and Total the Philippines: 1960

Fertility - 1980

Rate (TFR) in

Age-Group 1960a 1965a 197 0a 1975a 1980b15 - 19 84 74 56 56 4120 - 24 260 254 227 207 18025 - 29 313 313 302 248 21430 - 34 290 281 272 239 20735 - 39 211 216 199 168 15540 - 44 107 101 100 86 7745 - 49 27 20 22 27 22TFR 6.46 6.30 5.89 5.18 4.48a Figures show are average rates for 5-year period

wherein the year indicated is the midpoint. b Estimates based on projecting the 1978 RFPS values from

1978Source: NEDA,1983: Table II.7.

Urban-rural differentials in fertility were also observed to be evident in the country as a whole and each of the 13 regions. It was observed that women dwelling in urban areas have lower fertility than their rural counterparts. For instance in 1980, the rural total fertility rate for the country as a whole was 5.6 children while it was only 4 children per woman in the urban areas (Cabigon, n.d.). Education has been shown to be closely

33

(Cabigon, n.d.). Education has been shown to be closely related to fertility, and among the factors causing this fertility differential. Other factors that are related with fertility differentials are the use of contraception and probably infant mortality.

2.3.1 Education and FertilityFemale education has been demonstrated to have a

significant effect on fertility in many countries (Caldwell, 1980). However, the direction and strength of relationship of education and fertility is not consistent from country to country or even between regions within one country. In countries that have a higher level of education, education is observed to have an inverse relationship with fertility but in countries with low levels of education, women with a small amount of education tend to have a higher fertility than women with no schooling (Kandiah, 1981: 27).

In the Philippines, de Guzman's (1978) study of fertility differentials in the country, by considering women's education and residence, supports the inverse relationship between fertility and education - a pattern of declining fertility with increasing education. Further, the same study confirms the existence of urban-rural fertility differentials even after women's education has been controlled. In general, women in more urbanized area (Metro Manila) and other urban areas have lower fertility than women in rural areas. For instance, as seen in Table 2.2, women in Metro Manila aged 45-64 with 5-7 years of

34

schooling had 5.42 "other urban" and educational status, woman, respectively.

children, on average, compared with rural women of the same ages and who had 5.90 and 6.40 children per

Table 2.2 : Average Number of Children per Women aged 45-64 Years by Level of Education and Residence, Philippines:1973

Level of schooling Total MetroManila

OtherUrban

Rural

All levels 6.17 5.66 5.94 6.46College 1-4 years 4.56 4.37 4.44 5.04High school 1-4 years 6.00 5.37 6.04 6.51Grade 5-6 6.07 5.42 5.90 6.40Grade 1-4 6.61 6.65 6.35 6.68No schooling 5.95 6.54 6.17 5.84Source: deGuzman, 1978: Table 104 •

It is interesting to note however, as seen in Table2.2, and particularly in rural areas, that uneducated women have a lower average number of children per women than women with elementary education. Note also the reverse pattern of differentials in terms of place of residence: women with no education who live in Metro Manila have larger completed family size than women in other urban areas, and they have even larger completed family size than women in rural areas. Probably this is because women even with a small amount of education tend to shorten breastfeeding, and they are more prone to conceive and give birth successfully because of better health conditions, whereas women with no schooling (particularly

35

in rural areas) tend to be conservative and unaware of proper health and hygiene.

One of the more direct consequences of getting higher education is a postponement of marriage. Women who have higher education tend to marry later than less educated women. Engracia and Kim (1972) using the 1972 Fertility Survey in the Philippines, found that non-educated women marry about 4 years earlier than college-educated women. This length of time has a substantial effect on fertility in that it could mean an additional one child or two children to those who marry early. However, the same study further showed that differences in the timing of entry into sexual union do not sufficiently account for the differences in fertility level of various levels of educational attainment. Accordingly, while the delay in marriage is an important mechanism that brings significant reduction in fertility only among the less educated group, among groups with higher educational attainment, some other factors also influence the level of their fertility. Nevertheless, when level of education is held constant, age at marriage could explain part of the observed fertility differences between urban and rural areas (de Guzman, 1978: 135).

36

2.3.2 Contraceptive Use and FertilityBirth control methods are known to have been practiced

by women in the Philippines for several decades but the popular usage became apparent only after 1970 when family planning was launched officially in the country. To achieve the government goal of controlling population growth the family planning programme has embarked on three types of activities:

" First is education, to impress on the people the urgent need for controlling the population. The second is dissemination of knowledge on techniques of birth control sanctioned by scientific and medical practice. The third is the provision of facilities, especially in the rural areas, where assistance in the use of birth control methods may be extended to the poor." (Concepcion, 1978; 173)

The high level of contraceptive use among urban dwellers is often commonly advanced to explain the lower fertility in urban areas. The 1978 Republic of thePhilippines Fertility Survey (RPFS) reported 44 per cent of currently married fecund women in rural areas had neverused contraception and only 3 5 per cent are currentlyusing, as shown in Table 2.3. On the contrary, in urban areas, more than half (52 per cent) of the currentlymarried fecund women are currently using contraception while never users and had stopped using contraception were 24 per cent and 23 per cent respectively. Easy access in cities and urban centers to the means of contraceptive control probably explains the high percentage ofcontraceptive use among urban residents. In fact bothstudies of Cabigon (1978a) and Engracia et al. (1978)

37

confirm that accessibility of family planning clinics within the vicinity is highly related to acceptance and useof contraception Nevertheless, further analysis ofEngracia et al. (1978) employing multiple classificationanalysis shows, that fertility varies with theaccessibility of family planning services, but when thesocio-economic factors are controlled, most of therelationship disappears.

Table 2.3 : Percentage Distribution of Currently Married Fecund Women According to Use/Non-use of Contraception by Place of Residence, Philippines, 1978

ContraceptiveStatus

TypeUrban

of ResidenceRural

Never users 24.4 44.0Stoppers 23.4 21.3Currently using 52.2 34.7All women 100.0 100.0

(2575) (5317)Source: Cabigon, 1978a: Table 12.1.

The percentage distribution of ever-married women who had ever used any contraceptive method, by current age and family size, is given in Table 2.4. It can be seen that more women in age groups 25-34 and 35-44 had used contraception than women in other age groups. In terms of family size, women of family size 3 or more exhibit higher percentages of ever-use. This observation suggests that the practice of family planning is more common among the

38

older or middle ages women who probably have attained their desired family size. In 1973 and 1978, the reported desired family size among Filipino couples, as measured by the median, is 4.4 children. Though this is still high, it marked a reduction in the desired family size when compared with 5.6 desired family size in 1968 (UN, 1979: 9). In fact, the observed tendency among women who had already achieved their desired family size, was already observed since the early years after the official launching of family planning in the country. Engracia and Kim, using data of the 1972 Fertility Survey on Knowledge Attitudes and Practice of Family Planning, concluded "that use of family planning by Filipino women, is grounded on a motive to cut down the size of a family only after it has become disagreeably large." (1979: 9).

Table 2.4 : Percentage Distribution of Ever-married WomenWho Have Ever-used Any Contraceptive Method, by Current Age and Family Size, Philippines, 1978

Total15-24

Current Age 25-34 35-44 45-49

Percentage 57.5 46.9ever-used

66.1 60.6 37.5

Total0 1

5+Family

2size

3 4

Percen- 57.5 12.2 47.4tageever-used

59.2 68.4 65.5 59.2

Source: NCSO et al., 1979: Table 7.3.

39

2.3.3 Infant Mortality and FertilityPrevious studies have shown that infant mortality

experience affects a couple's fertility behaviour in four ways (see Taylor et al., 1976; Knodel, 1978; Preston, 1978) ;

(1) physiological effects in a situation where breast feeding is practiced, an infant/child death could shorten the duration of breastfeeding and amenorrhea, resulting in shorter birth intervals;

(2) child replacement effects - couples continue having children in order to replace those who die young, to achieve the number of surviving children considered to be sufficient;

(3) insurance effects- couples tends to adjust their fertility in anticipation of possible future deaths based on an awareness of the level of child mortality within their community; and

(4) societal effect - effects of infant/child mortality on •society operate indirectly through social customs to insure that fertility and mortality is brought into some sort of balance within the community.

Using data of the 1978 Republic of the Philippines Fertility Survey (RPFS), de Guzman (1978) seeks evidence to validate the physiological and child replacement effect of infant mortality on fertility. His study shows that women at the same parity level but who have fewer surviving children or have experienced high child mortality exhibited shorter birth intervals than their counterparts with more

40

survivors among their earlier births. Though the birth intervals difference between this two group of women is not statistically significant, he found that the larger differentials appear at earlier parities (parity 2 and 3), where the need to replace deceased children is believed to be urgent.

In addition, as Table 2.5 shows, the mean number of subsequent births is higher when a woman experiences a child loss. The more infant deaths, the higher the number of subsequent live births and the greater the difference observed between the means. The table also shows that rural women have a higher mean number of subsequent births than women in the urban areas even after parity and number of infant deaths are controlled. Differences in mean number of subsequent births between urban and rural are more pronounced at the first parity. However, rural-urban difference of the mean number of birth subsequent to parity 1 does not differ much whether woman have had infant deaths or not..

Taylor et al. (1976) pointed out, in their study validating the "child-survival hypothesis", that the greater effects of infant mortality on fertility operate through biological factors, that affect the lactational amenorrhea, and behavioural factors, whereby parents tend to replace their children who died at early age. However, as concluded by a number of researchers and demographers the observed higher fertility among women who experience infant mortality was not mainly due to the fact that parents tend to replace their dead children. Women who

41

experience child mortality can be a select group in several ways (Chowdhury et al., 1976; Balakrishma, 1978). They may belong to socio-economic groups which are likely or more likely to experience high mortality and high fertility rates.

Table 2.5 ; Mean Number of Birth Subsequent to Parity i by Experience of Infant Mortality at that Parityby Residence, Philippines: 1978

ParityNumberinfant

i /ofdeaths

Total Urban RuralRural-UrbanDifference

First parity0 4.27 3.68 4.54 0.861 5.08 4.48 5.37 0.89

Second parity0 3.87 3.34 4.10 0.781 4.54 4.09 4.68 0.592 5.33 * 5.45 -

Third parity0 3.57 3.17 3.73 0.561 4.01 3.42 4.17 0.562 4.97 4.75 5.02 0.273 * * * —

* Less than 15 casesSource; deGuzman, 1983a: Table 10.4

42

CHAPTER 3CHILDSPACING PATTERN AND DIFFERENTIALS

In the first part of this chapter, the methodology used to determine the pattern and differentials in childspacing will be described. An overview of thedependent variables, independent variables and the measurement used to carry out the analysis are presented. Specifically, life table techniques are employed to investigate the pace of subsequent fertility from parity 1 to 5 of the population by age, current place of residence, educational level, contraceptive use and experience of infant mortality. The results of the analysis will besummarized and presented in the succeeding parts.

3.1 The Methodology Used

DEPENDENT VARIABLES:The study considered the following dependent

variables:

Second birth interval: an interval from the first birth to either the second birth or the interview date.

Third birth interval: an interval from the second birth to either the third birth or the interview date.

Fourth birth interval: an interval from the third birth to either the fourth birth or the interview date.

Fifth birth interval: an interval from thefourth birth to either the fifth birth or the interview date.

Sixth birth interval: an interval from the fifth birth to either the sixth birth or the interview date.

43

It should be noted that open intervals can be included in this analysis because life table survival tecniques are used. In particular currently married women who are married only once and who have already reached parity n, whether they have attain parity n+l or not, will be considered in the calculation of the duration of months of each birth interval. To determine each birth interval, the month and year of birth of child n will be subtracted from the month and year of birth of child n+l. In cases of women who have not reached parity n+l, the interval after parity n will be obtained by subtracting the month and year of child birth n from the month and year of interview date. In cases of missing information on the month or/and year of childbirth, such births will be considered as missing cases and will be excluded in the analysis.

INDEPENDENT VARIABLES:

Place of Residence - Place of residence in the present study signifies whether the woman under study is currently residing in Metro Manila, other urban areas or rural areas at the time of the survey. The same concept of urban-rural definition used by the National Census and Statistics Office (NCSO) in undertaking the 1980 Census of Population was used in the 1983 National Demographic Survey (NDS). The urban-rural definition is given in Appendix - A.

Beset with the operational definition on place of residence, this study does not endeavour to establish directly the impact of modernization and urbanization on

44

the timing of subsequent births. Clearly, for the reason, a woman may have had part or all of her childbearing experience at other places which may be totally different from her current type of residence. Nevertheless, this study made a crude comparison of fertility behaviour between women residing currently in specific type of residence on the assumption that women of the same type of residence and specific characteristics have similar childbearing experience distinctive from the experience of women of other type of residence.

Out of the total 8,933 number of women under study, only 1,147 and 2,147 women came from Metro Manila and other urban areas respectively. Women from Metro Manila, other urban, and rural areas were expected to exhibit differences in the birth interval analysis but the smaller number of cases in Metro Manila and other urban areas in some instances, were insufficient for meaningful interpretation.

Education - Educational level in the present study refers to women's educational attainment at the time of the survey. Education in the country is either formal or nonformal. Formal education consist of three levels; elementary, secondary and tertiary or higher education. The number of years required to complete elementary and secondary levels are six and four years respectively, and the tertiary level requires at least four for an academic degree. However, instead of using the three levels, in

45

this study, educational level will be categorized into:(a) none/primary - women with at most 4 years formal

education(b) intermediate - women with 5 or 6 years formal

education, and(c) high school (HS) or college - women with more than

6 years formal education.

Table 3.1 shows differences in educational levels among currently married women who are married only once by place of residence. It can be seen that in Metro Manila the majority of the women have higher education. This is also the case in other urban areas to a lesser extent, but not in the rural areas. The proportion of Metro Manila women having high school or college education is 69 per cent compared with that of other urban and rural women of 55 and 32 percent respectively. It can also be seen in the table that younger women in all places have higher education than older women. This substantiates the better opportunity enjoyed by younger women to receive education, than in the past.

46

Table 3.1 : Percentage Distribution of Educational Level by Age of Women at the Time of Survey and Place of Residence, Philippines, 1983

EducationalLevel

A g e cif U o m e nTotal

15-24 25-34 35-49

X N X N

METRO MANILA

X N X N

none/primary 2.8 7 3.8 16 10.3 48 6.3 71intermediate 18.5 46 27.4 115 25.7 120 24.7 281HS or col lege 78.7 196 68.8 289 64.0 299 69.0 784

TOTAL 100.0 (249) 100.0 (420)

OTHER URBAN

100.0 (467) 100.0 (1136)

none/primary 10.8 45 9.0 72 18.8 169 13.5 286intermediate 28.2 118 29.1 234 35.4 318 31.6 670HS or col lege 61.0 255 61.9 497 45.8 411 54.9 1163

TOTAL 100.0 (418) 100.0 (803)

RURAL

100.0 (898) 100.0 (2119)

none/primary 18.4 227 19.6 391 34.9 816 25.7 1434intermediate 41.4 511 41.8 833 43.9 1028 42.6 2372HS or college 40.2 497 38.6 770 21.2 496 31.7 1763

TOTAL 100.0 (1235) 100.0 (1994) 100.0 (2340) 100.0 (5569)

Source: Computed from the 1983 National Demographic Survey (NDS) data tape

Contraceptive Use Status - The use of birth control methods during each interval was not asked in the 1983 NDS. Questions on contraceptive knowledge and practice asked of ever-married women 15-49 of age pertained only to women's status at the time of the survey. Though the previousinformation mentioned is more fitted to use in this birth interval analysis, it is not available, so the latter information will be used.

The 198 3 NDS data on whether or not a women had used contraception and was currently using contraceptioncontained the least inaccuraccy possible. The survey questionnaire consisted of some questions that could be

47

used as a counter check on the validity of previous questions, thus inconsistency could easily be re-checked during the field enumeration. Also, cases of intentionalmisreporting because of respondent's religious or personal reasons such as shame were negligible in number to affect over-all accuracy of the data.

Generally, women in Metro Manila and in other urban areas had better knowledge and access to modern methods of contraception than those from the rural areas. Table 3.2 shows differences in contraceptive use by place of residence. It can be seen that Metro Manila women have the highest percentage currently using contraception followed by women in the other urban areas with 53% and 44 % respectively. Of women in the rural areas, only one-third (31 %) are reported currently using contraception. Metro Manila women also exhibited the highest percentage using permanent method (ligation/vasectomy) and were the highest users of pills, IUD, and injection. Natural methods such as rhythm, withdrawal and abstinence were the most commonly used contraception methods in the country as a whole. In this study, examination of women by contraceptive use status will be divided into two groups: those who have ever used contraception which includes either currently using and not currently using contraception but have used before,and those who have never used.

48

Table 3.2 : Percentage Distribution of Currently Married Women, Married Only Once by Contraceptive Use and Place of Residence, Philippines,1983

P l a c e o f R e s i d e n c eT o t a l

Metro Manila Other Urban Rural

% N % N % N X N

Never Used 27.2 309 40.8 864 57.1 3181 49.4 4354

Ever Used 72.8 827 59.2 1253 42.9 2388 50.6 4468

not currently using 19.9 226 15.6 331 12.2 677 14.0 1234

currently using 52.9 601 43.5 921 30.7 1712 36.7 3234

Pill/IUD/Injection 14.3 162 8.9 188 6.7 371 8.2 721Ligation/Vasectomy 17.6 200 12.6 267 6.9 384 9.6 851Other modern method 1.4 16 2.5 52 2.2 122 2.2 190Natural method 14.3 163 14.8 314 11.4 636 12.6 1113Others 5.3 60 4.7 100 3.6 199 4.1 359

Total 100.0 (1136) 100.0 (2117) 100.0 (5569) 100.0 (8822)


Infant Mortality - The effect of infant mortality has been considered in this birth interval analysis by constructing attrition tables by birth order according to whether or not the previous child survived the first year of life. It should be stressed the fact that data used in this study on the number of children ever born and perhaps infant mortality is not free from errors. Cabigon's (1985) study of fertility and contraception using the same data set (1983 NDS) found a sudden fall in CEB at the last age- group, indicating an omission of children by the oldest age-group (45-49). Thus, analysis concerning women of this age-group should be taken caustiosly bearing in mind thislimitation.

49

Women in Metro Manila and other urban areas, with better economic conditions compared with rural areas, have lower percentages experiencing infant mortality. It can be seen in Table 3.3 that in Metro Manila the percentage of women experiencing infant deaths is just slightly lower than in other urban areas (9.2 % in Metro Manila and 10.8 % in other urban areas) but much lower if compared with that in rural areas (13.1 %) . Women at the younger age-groupthat is 15-24 years of age, have a lower percentage experiencing infants death than women at the older groups (25-34 and 35-49 of age). Perhaps this is because younger women have a relatively lower number of children ever born than the older women. It is worth noting too that the number of women who had experienced infant deaths is relatively lower than those who had not. Since a life table technique will be used in this analysis, the small number of cases in some of the categories can affect the results to some extent and thus one of the study'slimitations.

Table 3.3 : Percentage Distribution of Currently Married Women, Married Only Once Who Had Experienced Infant Death by Age Group and Place of Residence, Philippines, 1983

ExperienceStatus

A g e o f W o m e nTotal

15-24 25-34 35-49

X N X N

METRO MANILA

% N X N

experience 3.4 7 7.7 36 13.4 63 9.2 106did not 96.6 201 92.3 434 86.6 406 90.8 1041

TOTAL 100.0 (208) 100.0 (470)

OTHER URBAN

100.0 (469) 100.0 (1147)


TOTAL 100.0 (334) 100.0 (909)

RURAL

100.0 (905) 100.0 (2148)


TOTAL 100.0 (1045) 100.0 (2240) 100.0 (2354) 100.0 (5639)

Source: Computed from the 1983 the National Demographic Survey (NDS) data tape

51

SUMMARY MEASURES:This study focuses on two features of birth intervals:

the proportion of women at parity n who progress to parity n+1.

- the distribution of the timing of transition from one parity to the next for women who continue to have subsequent births.

Life tables will be calculated utilizing the procedure SURVIVAL of the SPSSX software package which will yield the life table functions needed in the analysis. The results will be presented by the life table birth function (Bx) in terms of the cumulative proportion of women having a subsequent birth at successive durations by birth order. To convey the results of the life tables in bulk without losing their values, the summary measures utilized by Rodriguez and Hobcraft (1980) are adopted:

B60 (Q) - the proportion of women having a subsequentbirth up to five years after the previous birth. Though the actual parity progression ratio is higher than the B60 and may need longer time than 5 years, B60 is considered to be a convenient indicator as five years is reasonbly long enough to provides a natural analog (Rodriguez and Hobcraft, 1980:12). This measure then is used as the indicator of fertility on the assumption that 5 year's experience is an adequate time to ascertain whether or not women at parity n will be progressing to parity n+1 or not.

52

Trimean (T) - trimean gives the average length of time (in months) that is taken to make a transition from the previous birth to the next birth. It is computed from the quartiles (q1; q2, q3) of the entire birth function (Bx) : Trimean (T) = (q-j_ + 2q2 + q3)/4. This specific measure had been considered in this study as a summary of the tempo or timing of fertility instead of the measure employed by Rodriguez and Hobcraft (1980) - the standardized trimean which is calculated based only from the birth function B60. The reason for the deviation made is that, if the standardized trimean is to be used it may give unrealistic and misleading results since it may underestimate in greater degree the average time elapsed before women have subsequent births. Though not shown in the present study, to avoid confusion, the standardized trimean based from the birth functions B60 and B72 were computed and been compared to the conventional trimean. The comparison shows that difference between the standardized and unstandardized trimean increases as parity progression ratio decreases. The difference between the two measure of trimean also vary for each different sub-group of women.

Median - The median is equal to the middle quartile (q2), which is the estimated durations (in months) by which half of the women in the sample have a subsequent birth. It will be sometimes be used as an indicator of the tempo of fertility in case the birth function B60 did not reach the upper quartile (q3 or 75 per cent) for the calculationof the trimean.

53

The average length of birth intervals will be presented in terms of the trimean and median in this analysis, but the discussion will focus more on the trimean since it is considered a more comprehensive measure than the median. Nevertheless, the median will be used in cases where the proportion of women having the subsequent births is between 50 per cent and 7 5 per cent, since the trimean cannot be calculated.

In the present study, the cutoff minimum number of cases for the life table analysis will be 50 cases, as Rodriguez and Hobcraft (1980) had used in their study. In some instances where the number of cases is between 20 and 50, the results should be treated with caution. In other words results from life table of cases less than 50 should be regarded as exploratory rather than definitive.

As mentioned in Chapter 1, this study partially retains selectivity biases and other biases because of small number of cases particularly at the higher parities. An approach suggested by Bumpass et al. (1982) cannot completely eliminate such biases by restricting the analysis to intervals initiated in the decade ending two years before the interview data. The selectivity bias has been examined in the first section of Chapter 4 of the present study, by controlling for age of women at giving birth, as suggested by Rodriguez and Hobcraft (1980:16).

54

3.2 Childspacinq Differentials bv Place of Residence

The birth function of women of parity 1 to 5 of the total sample and by different places of residence is presented in Appendix - C Figure 1. As can be observed in the figure, the proportion of women having a subsequent birth by each duration declines with parity in all areas. Selected duration of life table birth functions,summarized in Table 3.4, show that in Metro Manila, the proportion having a birth within 5 years (B60) falls from 90 per cent for parity 1 to 64 per cent for parity 5. The corresponding proportions in other urban areas are 91 per cent and 74 per cent, whereas in rural areas, the proportion of women having a birth within five years decreases gradually from 94 per cent for parity 1 to 83 per cent for parity 5. The pace of subsequent births among rural women differs with women in Metro Manila or with women in other urban areas. In rural areas, theproportions of women having second births to sixth births within five years show only a slight tendency to decline unlike in Metro Manila, the corresponding proportion of women having fourth to sixth births at the same duration are 71 per cent and 64 per cent, respectively.

55

Table 3.4 : Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Place of Residence, Philippines, 1983

B i r 1: h 0 r d e( r

Measures 2 3 4 5 6

METRO MANILA

B24 0.543 0.403 0.343 0.271 0.288B36 0.771 0.658 0.588 0.489 0.482B48 0.863 0.766 0.668 0.625 0.575B60 (Q) 0.901 0.809 0.708 0.703 0.639T rimean* 24.4 30.0 38.3 43.7 -Median* 22.9 27.4 30.4 36.5 37.4

N of cases** (1137) (937) (696) (465) (305)

OTHER URBAN

B24 0.487 0.391 0.323 0.293 0.307B36 0.770 0.697 0.609 0.584 0.582B48 0.870 0.811 0.729 0.701 0.696B60 (Q) 0.912 0.865 0.774 0.745 0.740Trimean* 25.2 28.7 33.7 34.7 36.6Median* 24.4 27.1 29.8 31.9 31.1

N of cases** (2118) (1779) (1411) (1022) (692)

B24 0.497

RURAL

0.381 0.334 0.311 0.303B36 0.811 0.726 0.674 0.651 0.644B48 0.906 0.846 0.798 0.771 0.766B60 (Q) 0.943 0.895 0.855 0.831 0.826Trimean* 24.7 28.0 30.1 31.4 31.8Median* 24.1 26.8 28.6 29.1 29.4

N of cases** (5566) (4890) (3996) (3112) (2323)


Notes :

B24 : the cummulative proportion of women who have had subsequent livelive birth within 24 months (2 years) after the previous live birth




* : presented in number of months.** : number of cases of birth intervals involved in

56

Generally, women in Metro Manila have lower fertility than women in the other urban areas, as shown by a smaller proportion of women having a second birth within 5 years and the faster decline in the proportion of women who continue to have subsequent births thereafter. However, Metro Manila fertility is relatively higher soon after the previous birth for the second to fourth births. For example, the proportion having the second births within 24 months after the first birth is 54 per cent in Metro Manila, while women in other urban and rural areas, the equivalent proportion of the same duration are 49 and 50 per cent. On the other hand, women in other urban areas exhibit consistently lower fertility than those women in the rural areas at all parities in all different durations, except at the third birth where the proportion of women having the third births within 24 months is slightly higher than that of the rural areas.

The trimean in each parity is also given in Table 3.4, to further compare the timing of births among women of different places of residence. As shown in the table, the average birth intervals in all areas increases sequentially from parity 1 to parity 5. Metro Manila women demonstrated the longest birth intervals at each parity (except the second birth) followed by women in other urban areas with average birth interval higher than that of the rural at all parities. Metro Manila women have the second births at 24 months while other urban and rural women have their second births at the intervals of 25 months on average. Similarly, at third births, the average birth interval

57

computed from women of different places of residence do not show much differences from each other. The slower pace of fertility among Metro Manila women is seen mainly after the third birth and becomes more pronounced at the fifth birth. The average fourth interval of Metro Manila, other urban and rural women are 3 8 months, 34 months and 3 0 months respectively. The trimean at sixth birth in Metro Manila is undefined since the birth functions does not reach the upper quartile. Nevertheless, the average birth intervals measured by the median shows women in Metro Manila have longer sixth birth intervals than the other two groups of women.

The longer birth intervals of Metro Manila women at higher parities and shorter to slightly higher intervals at lower parities is also reflected in the average birth intervals in terms of the median. Metro Manila women have second and third birth intervals of 23 and 27 months respectively, as compared with other urban and rural women, that have the corresponding second and third birth intervals of 24 and 27 months. It thus appears, based on the estimated proportion of women having a subsequent births and the average birth intervals measured by either the trimean or the median, that childspacing differences among women belonging to different places of residence differ only at higher birth orders, in particular after thethird births.

58

3.3 Educational Level and Childspacinq DifferentialsMore educated women are expected to have lower

fertility than the less educated women. This is visiblyobserved in Table 3.5 which shows the proportion of women of different educational levels having subsequent births within successive durations. Particularly at higher birth orders, women with higher education have a smaller proportion having subsequent births in each successive durations than those women with lower education. For instance, 77 per cent and 55 per cent of women with high school or college education, and 79 per cent and 65 per cent of women with at most primary education have their second and sixth births within 36 months. Within 60months, 72 per cent of women with high school or college education and 87 per cent of women with at most primary education have their fourth births. This may beinterpreted as indicating that women with higher education are likely to have fewer births.

Considering the average birth intervals measured by the trimean or the median, it is seen that the average birth intervals of all educational groups increases with parity. This is probably due to the fact that other factors including biological factors affects the length of birth intervals at higher parities. Women with the highest education have longer birth intervals in each birth order than the rest of the women, except with the second birth where differences in birth interval among women of different educational groups is not much.

Table 3.5 : Suimary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Educational Level, Philippines, 1983

Educational 8 i r t h O r d e r

2 3 4 5 6

none/primary B24 0.495 0.403 0.376 0.367 0.342B36 0.792 0.745 0.700 0.704 0.654B48 0.890 0.864 0.828 0.818 0.784B60 (Q) 0.932 0.914 0.879 0.869 0.831T rimean 24.9 27.4 28.3 28.5 30.7Median 24.1 26.5 27.2 26.8 28.6

N of cases (1791) (1652) (1478) (1273) (1047)

intermediate B24 0.506 0.392 0.321 0.285 0.287B36 0.828 0.738 0.683 0.623 0.623B48 0.915 0.863 0.805 0.748 0.752360 (Q) 0.944 0.905 0.857 0.810 0.815T rimean 24.4 27.7 30.1 32.9 32.8Median 23.9 26.5 28.6 30.1 30.3

N of cases (3321) (2977) (2500) (1989) (1462)

HS or college B24 0.497 0.369 0.313 0.265 0.277B36 0.768 0.659 0.566 0.526 0.553B48 0.871 0.765 0.667 0.646 0.629B60 (Q) 0.915 0.821 0.720 0.704 0.688Trimean 25.2 30.4 38.3 42.6 42.0Median 24.1 27.9 32.0 34.1 32.5

N of cases (3709) (2977) (2126) (1337) (811)


Note : see notes in Table 3.4

60

The proportion of women who have a subsequent birth classified by educational level and place of residence, is shown in Appendix - B Table 1 and Appendix - C Figure 2. It can be observed from the table and the figure, that in all areas, women with higher education tend to have lower fertility than less educated women. At a duration of 48 months (B48), in Metro Manila, the proportion of women with high school or college education having a fourth birth is 59 per cent while the proportion of women with at most only primary education is 84 per cent. The corresponding proportion among other urban women are 65 per cent and 80 per cent, while among rural women it is 71 and 83 per cent, respectively.

Quite noticeably, in all areas, the proportion of women having subsequent births in the lower birth orders do not show much difference between women of different educational groups. In fact, Metro Manila and other urban women with higher education tend to have their second births sooner after the previous birth than women in the rural areas of the same educational level. In Metro Manila 53 per cent of women with high school or college education and 50 per cent of women with at most primary education have their second birth within 24 months, and among women in other urban the proportions are 48 and 4 5 per cent respectively; whereas, in the rural setting, 49 per cent and 50 per cent of women belonging to the highest and lowest educational level respectively, have their second births within 24 months.

61

In addition, it can be seen that women with intermediate education exhibit higher proportions having subsequent births than those women with at most primary and with high school or college education, particularly at the early durations and lower birth orders. At a duration of 3 6 months, 82 per cent of Metro Manila women with intermediate education have their second births, as compared with the proportion of women with at most primary education and with high school or college education of 76 and 75 per cent. In rural areas, among women withintermediate education, 8 3 per cent of them have their second birth within 36 months while among both women with at most primary and with high school or college education the corresponding proportion is 79 per cent.

The negative effects of education on childbearing in all areas is further observed by looking at the average birth intervals as reflected by the trimean or the median shown in Table 3.6. The more women are educated the more they delay their subsequent birth. This relationship is particularly strong at the higher birth orders. However, in Metro Manila and other urban areas, the proportion of women having subsequent birth after the third birth among women with the highest education does not reach the upper quartile in the calculation of the trimean. Thus, comparison of average birth intervals will be done in terms of the median. In Metro Manila, women with high school or college education have their fourth birth on average at 34 months while women with at most primary education have a median of 27 months. Likewise, the corresponding intervals

62

in other urban are 32 and 27 months, and in rural areas are 31 and 27 months.

Comparing women of different places of residence with educational level being controlled, high school or college women in Metro Manila have longer birth intervals than the rest of the women of the same educational category, except at the second birth interval. Metro Manila women have their second and sixth births at 2 3 and 52 months, while women in other urban and rural have them both at 2 4 and 31 months. But among women with less education, Metro Manila women show shorter to only slightly higher intervals than those women in the other urban or rural areas. However, these results should be considered with caution since the sample number of live births from the less educated women in Metro Manila from which the intervals are calculated is relatively small, particularly in the none/primary category. Nevertheless, in general, it appears that effects of education in delaying the timing of subsequent births at the higher parities became substantially strong only when women with higher education are involved. This is probably due to the fact that other factors in relation to higher education compounded the effects of education on the timing of childbearing.

3.6

:

Sunn

ary

Mea

sure

s fo

r B

irth

In

teva

ls o

f B

irth

ord

er 2

,3,4

,5,6

by

Edu

catio

nal

Leve

l an

d P

lace

of

Res

iden

ce,

Ph

ilip

pin

es,

198!

LA3 s 28

.3 CO3

3OO 32.0

29.6 o

oos 35

.631

.3

o o-

o

jO o 4 nT KlKl oo vO O LAoo 00AJ 3AJ 5 CO CMKl oAl Kl4 r ! Aj Kl Klo O w o

3

28.2

27.2 o £ O O 3 3 LA CO

CO N. 0O 3 3AJ o N- 4 O Kl Klo-

o-

O

N- CM LA o O CO 4 •4 AJ5 n:CM 3AJ

O-ki

o n:CM 3AJoKl 3 o a!AJ AJo

-o CM o

ft in o 4 KlLA in 4

2369

) COKl O AJO'o CM CM Kl4 O 4 • AJ O 3 O

oLA -4 Al AJ

o N" 00 o• Kl

o ^• LAO' LAAl »—

"O LA O ^00 IM O O

■ k> (M r\jO w

-4 O' AJ3 CO N-• (M fM

CM PO 00kiO' LA 4 • fM fM

AJ-O

o oKl00 CM • Kl

«5 O KlX AI> 00 • fM

o <5

S z

<0 COo c3 O

1111 >•

> 1 ^ i4) ■—2 Z

a> • • O Q. C

KlCM

O • •O' CO N- 'O• CM CM CM

ä

3 ^ °N 'O Kl • Kl Klo5

^ N* >os in n cO• Kl Kl KlO w

28

AJ O' 4 ^Kl 0000 O O' O'■ IO AI vf

CO -O -4S N- 3• fM fM O

O 'O «- AJ O'OO' >4 4 S• AJ AJ -OO w

o'O

vO O' s- AJO O'• AJO

S. N- IA O'OO LA 4 ■ AJ AJ

*CM >4 oo NO CMrsJKl

in s inKl AlKl S CMino o

LA

a)2in o

57.6 O

oj >os-8

<*o

o’ o

26.8

(63) O1 31.0

26.7 3

33>o 34

.5

o w o

r-8

oo 'O 3 8 Kl «—LAAl 2 inCM KlAJ CM N. ~4 O Kl Klo’ w O

24.1

(71) oO' 23.0

22.1 CM00CM

KlO'OO 24.7

23.2

o w o

(Aa> (/Io</)CD c (A(0<0 c4> <0s o o 104» s o a

3ec_ s

£ — Z 2X z CO i— X z CO 1- X

O <voi1 (0 Lm 41CJ

1o

~oc £ 00 o

in

O'AJki

•4 AAJ N.K> sO

O

A JO

fM

CMCMvO

8.2

2“O

$>

Io

’■£.CD

(A *

•M

DM3

;ee

note

s in

Tab

le 3

.4h

is c

ell

sh

ould

be

trea

ted

with

cau

tio

n d

ue t

o s

mal

l sa

mpl

e si

z*

64

3.4 Contraceptive Use and Childspacincr DifferentialsIt is worth mentioning beforehand that "ever-user" and

"never-user" of contraception, as defined in this study, refers to women's contraceptive experience at the time of the survey and do not refer to any specific birth intervals. Consequently, the specific interval when a woman first used contraception and the length of period she had used it are not known. Though contraceptive use has long been recognized to play a vital role in explaining fertility differentials, because of the above limitation this analysis will only give the most likely effects of contraception on fertility behaviour among women of different places of residence - Metro Manila, other urban and rural areas.

Table 3.7 gives the overall picture of fertility behaviour of women by contraceptive status. It can be seen, at second birth order, women who have ever used contraception have a higher proportion having subsequent births than those women who had never used contraception, but at higher birth order the reverse is observed. After the second births, a smaller proportion of women who had ever used contraception continues to have subsequent births than women who had never used contraception. Ninety-five per cent of women who had used contraception and ninety-one per cent of women who had never used contraception have their second births within 60 months. The proportion of women having their third and sixth births, among women who

had ever used contraception is 87 and respectively, and among never-users it is cent correspondingly.

65

73 per cent 89 and 84 per

Table 3.7 : Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Contraceptive Use, Philippines, 1983

Contraceptive B i r t h O r d e r

2 3 4 5 6

Ever Used B24 0.535 0.387 0.330 0.283 0.293B36 0.818 0.700 0.628 0.591 0.572B48 0.911 0.815 0.736 0.709 0.673B60 (Q) 0.949 0.867 0.791 0.761 0.731Trimean 23.8 28.5 32.8 35.5 37.8Median 23.1 26.8 29.4 30.7 31.6

N of cases (4466) (4026) (3236) (2379) (1632)

Never Used B24 0.562 0.385 0.335 0.324 0.312B36 0.774 0.723 0.674 0.650 0.659B48 0.873 0.843 0.804 0.773 0.791B60 (Q) 0.910 0.890 0.853 0.837 0.845Trimean 25.8 29.2 29.9 31.4 30.9Median 25.0 27.0 28.5 29.2 29.0

N of cases (4352) (3578) (2864) (2219) (1688)



The average birth intervals measured by the trimean shows that ever-users tends to have shorter intervals at the second and third births. Ever-users have a trimean of 24 months for second birth intervals while never-users have their second child at 26 months on average. Indication of longer birth intervals among ever-users is seen only after the third births. Women who had ever used contraception

66

have their fourth and sixth births at 3 3 and 3 8 months while women who had never used contraception have them at the intervals of 3 0 and 31 months. Note that "ever-users" in this study were based on the woman's experience status at the time of survey and does not relate to a specific birth period when did woman started to use contraception. Probably, this evidence shows that most "ever-users" practiced contraception after having their second birth. However, if "ever-users" used contraception at early parities, probably the preference for more effective contraception in spacing birth is not as great as to limit birth after having 3 children.

The proportions having a birth within each successive duration for women who have used and those who have never used contraception, from parity 1 to parity 6, are shown in Appendix - B Table 2 and Appendix - C Figure 3. It can be observed that in Metro Manila, the proportion having subsequent births among ever-users is higher than for never-users in the second and third births in all successive durations. In other urban areas, the proportion of ever-users having subsequent births is only higher than never-users at second births, and in third births at the early duration (B24). In rural areas, the proportion having subsequent births among women who had ever used contraception is only higher at the second births. In Metro Manila, 56 per cent of ever-users and 50 per cent never-users have their second births within 24 months while in other urban the corresponding proportion are 52 and 43

67

per cent, and in rural areas 53 and 4 6 per cent respectively.

Generally, after the third births, the proportion having subsequent births among ever-users dramatically decreases. In contrast, among never-users, the proportion having subsequent births from the fourth to the sixth births displays only a slight tendency to decline. The observation is particularly pronounced among Metro Manila women. At a duration of 60 months, the proportion of Metro Manila having the fourth and the sixth births among ever- users is 8 3 and 77 per cent while among never-users the proportion is 88 and 87 per cent. Perhaps, because up to four children is the most preferred number of children among Filipino women, the use of contraceptives was delayed in the lower birth orders.

The average birth intervals in terms of the trimean and the median among women from different places of residence by contraceptive use are given in Table 3.8. It can be observed in all areas, that at lower birth orders women who had ever used contraception have shorter birth intervals than those women who had never used. But at higher births, that is from fourth births thereafter, the reverse is shown - longer birth intervals among ever-users than among never-users.

Since in Metro Manila, the proportion of subsequent births within five years did not reach the upper quartile, the median will be used to convey to the average birth intervals. Ever-users in Metro Manila have their second births in 23 months, compared with 24 among never-users.

68

Ever-users in the other urban and rural areas have their second births also in 23 months while among never-users, women in other urban have their second births in 26 months and rural women in 25 months. Within intervals of 31 and 38 months, ever-users in Metro Manila have their fourth and sixth births respectively, while never-users in the same area have them in 28 and 37 months. The corresponding proportion among ever-users in the other urban areas is 30 and 32 months, and in the rural areas it is 29 and 31 months. Never-users in the urban areas have their fourth and sixth births at 29 and 30 months respectively while it is 28 and 28 months from rural areas. This evidence further confirms that fertility behaviour differentials is more pronounced at the higher parities. It further shows that, Metro Manila women still have higher birth intervals than the rest of the women, even after contraceptive usehas been controlled.

eo s avO IM N- N- £

0.86

5930

.19

28.5

3

§o • in o O Kl Kl w ro

Kl CO K- vf OX-N co N- in in ro • X

(.in O • OS N IM ■ Kl O

co • oo• O CMo ro E

a>TDu

O vf Rur

al

0.82

9431

.128

.78

(181

1)

0.87

6329

.65

28.3

9

(218

5)

JC•ML. Kl .8

863

28.3

26.9

1

oCMCM 0.90

327

.76

26.7

1

SCMs *•" o wo c a

inVr^ ro xdS ' 0. - : §

o- o in vf in oo ca oN-c (M O Kl Kl ■ cm ca roCM O in vf ■ ca ca K1

o w o w

£ N a s s! '"I £ w a s o ab Ki s CM

o 00 • O Kl O Kl C~ Kl O ■ Kl Kl CMw“O O

CO<DQC in Ovf ca s o ca co vr in Kl

(383

)

*♦- in ca i K- cao M

roXDw fc * ~• Kl Kl

U 0)

o. “O s O O eoO C~ Kl X eo in KlC*. N. »- £'S

L.o Nt O

b~ v f o • Kl Klo00

OS- r— o • Kl ca inw(0 o5 o o<uin -C 53 4-» o Kl 00 O » - ( M O ’ o

00 00 v fO V f o O> L. Kl 28 « XD • C\i CM 00 O 00• ca ca

N-' S

O o.

Om £ fc 0O Vf*— ca

(99)v f 03 S ‘ roCM £ ‘ kro o w oca

V•O

.698

9

36.0

2 roXDO •“»- KlCv . o

o i n ro . OO Kl 35

L. oL. <D <0CO ■o S- ON- O v f OO«- o Kl

o v f z £ ’ ^ • Klroinw

ca • •N- 00■ cavO

(A o O o

> JC 52 4-» £ CO K l r -

ca v f ca S<- (M (M O O N .

c L. Kl co i nO ^ o rv CM

5 o d£_ CO

aü S- O c- o o m rv <o o 'O in o oo*♦- ca c> v f ca ^ (M (M

CMCO 3 i n Kl ^ CM CMg

(A<D

CA <0Z

C/)V

(A<V

in<uu *■* Ca <o cw g (0

C/)(0ü^ c o ca cw ^ CO

(AxüO

>> in10 >

00ro o cn u

(A

'Si

0) 2 %3

Jrs

•Mt n aJ 0)>o £ oz

8.

2<ö_g

%>

10

£ .2o>1o

co

’SzrO00O

'S

CT»VD

70

3.5 Infant Mortality and Childspacinq DifferentialsThe "child-survival hypothesis" argues that improved

child survival could contribute to increased family planning motivation and thus to fertility decline. From a biological point of view, a child death during infancy could shorten duration of breastfeeding and consequently post-partum amenorrhea resulting in shorter birth intervals. Likewise, it could shorten the post-partum sexual abstinence. From a psychological point of view, there is the tendency among couples to replace their lost child during infancy much sooner, resulting in shortened gaps between births. Although several studies had given evidence empirically about the effects of infant mortality on fertility, still it remains debatable and open for criticism.

Table 3.9 shows the proportion of women having subsequent births at each successive duration and birth intervals, at specific parities, by survivorship of previous live birth in the first year of life. In general, women whose child prior to the specific birth survives in the first year of life, have a lower proportion having subsequent births in each successive duration and longer birth intervals than those women whose previous child died during infancy. At second and sixth births, 49 and 29 per cent of women who did not experience prior infant deaths have subsequent births within 24 months, as compared with the proportion among women who had experienced infant deaths of 76 and 59 per cent respectively.

71

Table 3.9 : Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Survivorship of Previous Live Birth in the First Year of Life, Philippines, 1983

Survivor-hip B i r t h O r d e r

2 3 4 5 6

Survive B24 0.488 0.248 0.318 0.291 0.287B36 0.792 0.706 0.641 0.614 0.612B48 0.890 0.825 0.762 0.738 0.732B60 (Q) 0.929 0.876 0.815 0.796 0.732Trimean 25.1 28.7 31.7 33.4 33.8Median 24.3 27.2 29.4 30.3 30.5

N of cases (8439) (7250) (5832) (4388) (3151)

Did not B24 0.762 62.010 0.622 0.514 0.594survive B36 0.889 0.820 0.821 0.708 0.723

B48 0.940 0.899 0.893 0.785 0.779B60 (Q) 0.964 0.921 0.916 0.843 0.824Trimean 18.0 21.0 21.3 26.3 24.5Median 17.6 20.3 21.0 23.2 21.9

N of cases (378) (341) (259) (197) (160)



The trimean difference between the two groups of women according to infant mortality experience ranges from 7 to 11 months. Women who did not experience death of first child have on average 25 months second birth intervals and women who did not experienced deaths of the third child have on average 32 months fourth birth intervals. Women who did experience infant deaths at the first and third births have second and fourth intervals of 18 and 21 months. While this result shows the expected pattern, according to the child survival hypothesis, it should be taken cautiously because of the relatively small number of

72

cases of women whose previous live birth did not survive in the first year of life. Nevertheless, it is quite obvious from the table, that there is a tendency of faster pace of subsequent births among women whose previous child died during infancy, particularly at parity 1 to parity 3. Ninety-six per cent of women whose first child died during infancy have their births within 5 years and 92 per cent of women who lost their third child during infancy have fourth births. The corresponding proportion among women who did not experience such infant deaths is 9 3 and 82 per cent. Possibly, at lower parities, the need to replace deceased children is more urgent.

Effects of infant mortality on the rate of childbearing among women of different places of residence are shown in Appendix - B Tables 3 and Appendix - C Figures 4. It can be seen that in all areas the proportion of women who had experienced prior infant deaths in each successive duration and birth order is higher than with women who did not experience prior infant deaths. This is particularly true at lower birth orders. Although Metro Manila women who had experienced prior infant deaths tend to have their births sooner after the previous births, particularly at second births, their pace of subsequent births decreases faster than those women in the other urban and rural areas at the third births and thereafter. Eighty-one per cent of Metro Manila women whose first child died at infancy have their second births within 24 months while 4 6 per cent of women who had lost their second child at infancy have their births at the same duration. In

73

other urban areas, the equivalent proportions of women are 66 per cent and 69 per cent, and in rural areas the corresponding proportions of women are 78 per cent and 62 per cent. However, results from Metro Manila and other urban areas under the category "did not survive" should be treated with caution since these figures are based on a small sample of cases. This problem limits further comparison among rural women and other urban women who did not experience infant deaths.

As shown in Table 3.10, in rural areas, the faster pace of subsequent births among women who had experienced infant deaths can be observed. Among rural women who had experienced loss of first child at infancy, 96 per cent of them have their second births within 60 months while for those women who had lost their fifth birth, 88 per cent of them have sixth birth at the same duration. Among rural women who did not experience prior infant deaths, the corresponding proportion within the same duration are 94 per cent and 82 per cent. The average birth intervals in terms of the trimean shows that the presence of infant deaths in the onset of the interval tends to shorten the gap between live births. Rural women who had experienced prior infant deaths have their second and fourth births at 18 and 21 months while women who did not experience prior infant deaths have their second and fourth births at 25 and 31 months. The table also shows that among women who did not experience infant deaths Metro Manila women have longer birth intervals at higher parities compared with their counterpart women in other urban and rural areas.

3.10

:

Sum

mar

y M

easu

res

for

Bir

th

Inte

vals

of

Bir

th O

rder

2,3

,4,5

,6 b

y S

urvi

vors

hip

of P

revi

ous

Live

Bir

th

in t

he F

irs

and

Plac

e of

Res

iden

ce,

Ph

ilip

pin

es,

1983

(54)

O

a t

fM

in f\J N PO "O POrg00 CM O

• ro cmoCM

§ d d■ fM fMo

o CM O w

.82

9

31

.8

29

.4 so .8

57

23

.3

21

.0

o CMw o w

23 .8

51

30

.6

21

.0 O5 .9

34

20

.8

20

.2 5h-GC o Kl o

.89

5

28

.4

27

.1

PO «- v* f\J • •

/-xin«o

o * - o• fM fMCMCM

o vj- O x^.9

42

25

.0

24

.4 S3 .9

60

17

.8

17

.3

s<M

o inw o

o -$• PO o•r - i t— 00in R i fM

fMPO

■ poo S • CM

OX

0.7

41

32

.0

(98

3)

0.8

33

31

.0

27

.5

(36

)c

&L.

).7

66

34

.6

30

.4 ininro ).

95

7

19

.4

18

.9

(52

)

«5w

o

.85

8

29

.2

27

.5 inOvO .9

81

19

.2

17

.9

(81

)

o r—w

o

O ooO in vt

roN- N. COrÔ' O' N.

■x£

• (M fM O 1 d *” *~ w

8 .2

CO COPO

O COPO oo

2■§O i -O

• PO00CM

'O ■ «-■ in c

QZ

).7

00

36

.6

/-x

5.7

32

48

.8

(16

)

>

(0 (j

o °°/•x o

fM oo -ira ” i dd f° 1

N- • -Od ^

CMX«/

Q.(0

o O)

2 SO fs» Q

5 d• PO fM

o w

OCO »— vj-

• PO fM O

rorow co

ra00 S- fM O' • •

■x CM O Oo o

z00 N* ro• CM CM g O ' CO O ' x *

WPOCO

Ow

o O'

0>C/>

<D(/)

-C

B60

(Q

)

Tri

me

an

Me

dia

n O

Z B60

(Q

)

T ri

me

an

Me

dia

n O

z

§L.

134-»

<uo

>v > o • - c > 2

>L

CO

T3 3•— (/)Q

C-

iCO

75

CHAPTER 4EFFECT OF COHORT DIFFERENCES ON CHILDSPACING

This chapter examines the effect of cohort differences in childspacing by controlling for women's age at the time of the survey. The two cohorts of women to be compared in this Chapter are, first, those women who were aged between 15 and 34 at the time of the survey, and second, those women who were aged between 3 5 and 49. The first section of this chapter considers cohort effects in general for all women in the sample.

Effects of selectivity biases on cohort differences are also examined in this chapter by constructing separate life tables with age of women at giving birth being controlled. Age at giving birth is divided into two groups: younger age and older age. Younger age at giving birth refers to women who had second and third births at ages less than 25 and those who had fourth to sixth births at ages less than 30. Older age at giving birth refers to women who had second and third births at age 25 and over and those who had fourth to sixth births at age 3 0 and over.

The effects of education, contraceptive use and infant mortality on childspacing controlling cohort differences are examined in the next three sections of this chapter. The small number of cases in Metro Manila and other urban areas when age is controlled give insufficient basis for the calculation of the life tables. Thus, further

76

comparison of childspacing practices among women of different age cohorts by place of residence was not covered in this analysis.

4.1 Cohort Differences in ChildspacingSelected successive durations in Table 4.1 shows that,

except at parity 1 and parity 5, the proportion of younger women having subsequent births in all successive durations is lower than those of the older women. Eighty-seven per cent of younger women of parity 2 have third births within 60 months and 77 per cent of these continue to have fifth births. The proportion of older women having third and fifth births are 88 per cent and 81 per cent, respectively.

The pattern of increasing average birth intervals as parity increases can also be observed in the two groups of women, except at parity 5 among younger cohorts where intervals measured by the trimean are shorter by 4 months compared with the fourth birth intervals. In fact, at the sixth birth order, younger cohorts have shorter birth intervals than the older cohorts. However, though the difference is not much, it can be observed that in all other birth orders, younger cohorts space their children longer than the older cohorts. Younger women have their second and fifth births at 28 and 34 months while older cohorts have them in 25 and 3 2 months. The average sixth birth interval of younger women is 30 months while that of older women is 34 months. Possibly, the younger cohorts with six or more births are selected groups in that they have closely spaced births to belong in that category.

77

In common with other childspacing studies, the sample of this study is selected only for currently married women who have already reached parity n at the time of the survey. Investigation of selectivity biases and its effects on cohort differences is thus necessary to be aware of the reliability of the results. This is particularly true for results at higher birth order of the younger cohorts, where there is a higher possibility that the selected younger women may have a higher fecundity and may not represent the whole population. To examine selectivity biases, cohort effects on childspacing are re-examined with women's age at giving birth being controlled.

Table 4.1 : Summary Measures for Birth Intervals of Birth Order 2,3,4,5,6 by Age at the Time of the Survey, Philippines, 1983

Age at Time of Survey

B i r t h O r d e r

2 3 4 5 6

15 - 34 : B24 0.509 0.369 0.324 0.274 0.335B36 0.799 0.688 0.626 0.604 0.674B48 0.895 0.809 0.744 0.717 0.796B60 (Q) 0.932 0.872 0.805 0.772 0.844T rimean 27.7 29.0 32.7 34.5 30.1Median 23.8 27.6 29.9 30.2 27.8

N of cases (5114) (4020) (2766) (1677) (882)

35 - 49 : B24 0.489 0.401 0.339 0.316 0.293B36 0.793 0.731 0.664 0.626 0.602B48 0.890 0.844 0.782 0.748 0.720B60 (Q) 0.929 0.884 0.829 0.807 0.779Trimean 25.0 27.8 30.7 32.5 34.5Median 24.3 26.4 28.3 29.7 30.9

N of cases (3706) (3586) (3337) (2921) (2438)


Note see notes in Table 3.4

78

As shown in Table 4.2, even after age at giving birth is controlled, women aged 15 to 34 at the time of the survey have a lower proportion having subsequent births in all birth orders, compared to those women aged 34 and over at the time of the survey. Among women giving birth at younger ages, the second and sixth birth intervals among younger and older cohorts do not vary, but among women giving birth at older ages, younger cohorts exhibit shorter second and sixth birth intervals than the older cohorts. However, in the third and fifth births, younger cohorts seems to have longer birth intervals than older cohorts in either younger or older age at giving birth. This provides the evidence that younger cohorts who have their sixth births at older age tend to be affected by selectivity, in that for women to be in this category they must have shorter intervals. On the other hand, the observed shorter intervals among younger cohorts having second births at older ages may not be a spurious result of selectivity but rather suggest a recent change on fertility behaviour among younger cohorts - a pattern of more rapid subsequent second births after the first birth among younger cohorts.

Having observed the probable effects of selectivity biases only at the sixth births, further examination of cohort differences on childspacing by socio-economic characteristics of the women will be carried out without controlling age of women at giving birth. Nevertheless, all further analysis will check for possible selectivity biases that may be present, especially at higher parities.

le 4.

2 :

Summa

ry Me

asur

es f

or Bir

th Int

ervals

of

Birth

Order

2,3,

4,5,6

by A

ge at

the T

imeCo

ntroll

ing b

y Age

at

Givi

ng B

irth,

Philip

pines,

1983

>I

JC4~*

JQa<-> 'o >-<0 01a l l< I— CO

t

aL.

c

O)s

<do>(0«o

09 00 » O « -N- in "O 'O •m 'O 'S 'S o>............CMo o o o

«— O' N- -O«— m *— •<- m m vf......................................- t-o o o

O' U1 *— t— ON- •■f CM CM •ro in in o o o o

(M O 'O N- o eoin o t* O' • •n o s s in oo■ ■ • • no cmo o o o

(M st in oo vj inco no o no>t S- s eo >0 >t• • • • (M CMO O O O

3

o o o o

O O 00 » - S MO CM no Ocm K cm Os• • • • no cmo o o o

in m o s— io N-iM in in i- ■ •in >o n «O w a• • • • no cmo o o o

£C_

a

CT>COCD *- no sO O CM CMo) co in o • •<o m n oo co oo s• • • • CM CM0- o o o o

o o o o

o S cw C£ CD

•~t VO 00 O • - -5cm no vt so c- <ooo q co 03 i

o

£' O

s o o s— ^no o cm N- ■ • inin s oo co co s cmCM CM sO

uoono

ro o o o st n- **o o st ■ • nin oo o o st ro to• • • • CM CM «—

-■o m in co noin i- m o •cm in 'O 'O in......................................mo o o o

8

00 st o> ^ *■»m co 0> sA . ocm vt m >o n- -fo o o o

«— eo sO o «—■O S - O s tcm vt m sO co

• • • • I noo o o o

S r- CM sf N <-sO in N- CM • •no "O N- oo o co• • • • no cmo o o o

r*- o o s- o5 R 5 S s ^• • ■ • CM CMO O O O

s t O cm i n s t-o *— ro eg • •no N- oo oo co S-• • • • CM CM

O O O O

O CM M <- S CMm T- m eg • •in N co co co N• • • • CM CMO O O O

CM O' sO O' CM O-O cm st egtn N- co co eo N-• • • • CM CMo o o o

0 so cm no ro socm 0- o no«f s- O' O' 'O in• • • • CM CMo o o o

in O' O' *— -f O'o »— «- in • •in co O' O' 'f no• • • • CM CMo o o o

cO <0 c^ CD

t <0 CO O - Dcm no st so «- 0)co aa co co i— x

5§

?o

8.22“S

%><1o

2

o’SzrocoO'

’S

T32 *

</>

'S o(D 4-»

I fU D 01 O(/> H- 4-» 4-f<0 co

0) oo o? 'SCD CD

in o cm no

O) O)

— c_

I f5 *

'S ao «-0 oCD O in s»-*-> *■>CD CDin o cm no§ §a ain in in tn01 o>

<D a>a> a»CD CD

a o

80

4.2 Educational Effects on ChildspacinqMore educated women have lower fertility than less

educated women, a preposition supported by results given in the previous chapter showing a smaller proportion having subsequent births, and longer birth intervals. Nevertheless, the relationship of education and the pace of fertility among Filipino women is examined again in this section with women's age being controlled to ascertain whether or not differences in the rate of childbearing among women of different educational groups is simply a cohort effect.

It is anticipated that younger women, who tend to have more education than older women, have a slower pace of fertility. This expectation is probably true, as results in Table 4.3 and Appendix - C Figure 5 show. Except at the sixth birth which, is largely affected by selectivity bias at higher birth order, younger cohorts with at least intermediate education show a general tendency towards a smaller proportion having subsequent births at higher birth orders than older women of the same educational level. But among women with at most primary educationonly, the proportion of younger women having subsequent births does not show much difference nor any specific general pattern.

The inverse relationship of education and the pace of fertility at higher parities remains within each cohort. After second births, the proportion of less educated women having subsequent births is higher than that of more

81

educated women at each duration and birth order in either younger or older cohorts. For example, among younger cohorts, with at most primary education, 92 and 88 per cent of women have third and fifth births within 5 years while among those with high school or college education the corresponding proportions are 83 and 66 per cent. This pattern is also evident among older cohorts where the proportion of women with least education having third and fifth births within 5 years are 91 and 87 per cent whereas with highest education, the proportion are 82 and 72 per cent. It can be concluded further from these results that, although negative effects of education on the timing of childbearing exist within each cohort, effects of higher education on the pace of subsequent births at higher parities was more visible among younger cohorts than older cohorts.

The average birth intervals in terms of the trimean, as shown on the same table, indicates that younger women with at least an intermediate education have longer birth intervals after the second birth than the older women with the same educational level. Conversely, among women with at most primary education, the birth intervals from the younger cohort do not differ from those of the oldercohorts.

ca

oX*.wvi_tI)CD

X>.

POxl0)

2«J

to01

to«X

■M 'o >sto V« g £O) — D< > — (/)

O § pi in xl xl CM co CM px. in vf -•-o K» in NO inPO drg roCM CM in 3 xO ro roro inO o o o w O o d d

L. ininroCMc~y 3.

517

3.62

53.

662 xl

POO

POroinin

CMCOCMCMroin SNÖ

Pxiv

ino

inroro

rocor^

1983 <D <DO)

O o o o

C/)0>L.o xl 2 Oro

rgin

fx.3

CM£

00ChPO POPO

✓-NtZro

in00inro3

oCMN-oPx’PO

o

ro ooc

1 JC oo 6 o' o o o o O

£

+-»L.

po

v*roo 3.

634 rginfx

o 3.82

9 Px

E

oCPrg

POPO00

oNt ! o

s sCOrgCPrg

O>6CM

Nt

f CD W o

uCO roco 3 rC £ in in V*- CMCM s

0.86

7 o Px V* inrg >*■

ofx.O

006

o6

inCM rg oinCMino

Px6

§o rg roCM CM

jH—

•M O 3.31

03.

683

COo 3.

890 o

3

POco’rg

/•xo

3.28

0 oxOinPC 5OO

OxPOPO

O

3CMNOo

<uo>< inPCCM

xlE 5Px

in

PO

in

3cohi

§rg xOCPinfx.

Ox5

mrgPO

o

31? t- O 6 6 6 w 6 d d dto <D

T3 2>0> L.

o xlto

1 3.32

63.

670

3.78

8 in00

fv

3rgCPrg

in coPO oixO

in5

PO ooCPrg

POcdrg

ininro

co45 w o o o o C

-C •M

%u

•wL. PO 3.

383

3.71

6 xloSo

inCP00o

oco’rg

cov6rg

CMininoxl

oinPx

jx-00

xlCP

POpjCM

PO»o'rg

inCMv£2 CO

£ 3 vOPO o s O PO va co oo o OxPO O xOo rg in CO £ o POrg POrg in00 xl CO Six Ox inCM NOin o o 6 6 o d o d T—xlpo”

rg*

«-o ino o>rg rgPO oin - xl o OOrg xOPO nj POrg in xl COoN.t-o vO ro Px oo co COrg ■org inCM PO -o Px co ro CPrg

-Co o 6 6 w o d d d w

*->!_co f\j NOo noPO fc O O § xl .— 3 in CO

in ro o 00 00 00rg ■Org 5PO 00 co COCM xOrg OO

to L. o o’ 6 o o d d d

<2 N . 00 O x l p . 00

— m 3 eo co aJC_ • • • • rg rga. o o o o

CO N M O > f S^ ^ lO C\J 1 ■po Px co Ox fx x<j■ • • • rg (Mo o o o

t > g IO 1 - Q NPO «— CP OJin oo oo o> vj- po

• • • ■ cm njo o o o

cO to cw ^ (0

g <o co o - 'p(M M >0 C- 4>oo cd co m »- x

inPO-o

£

in in po o «-oo o ro egfO N. oo oo oo

• ■ • • n jo o o o

in xi eg «- xi xiO xl xn .— • •xi s- oo o Px vo• • • ■ eg rgo o o o

-n n. rx- -o in o-O s- og mg S oo o in g• • • • rg rgo o o o

O <5 Cw fflg «o oo o - prg po xi o u £co co co ao i— x

pi<>

/•“N

5

8.

2

2"D

£>

1u

’■s.2

CMCO

83

The observed tendency for more educated women to postpone their subsequent births after the second births, as found earlier, still remain within each cohort. Younger cohorts with high school or college education have a trimean for their third and fifth birth intervals of 32 and 43 months while the corresponding intervals among women with no education or primary education are 27 and 29 months. Older women with the highest educational level have their third and fifth births at 29 and 41 months while those with the least education have third and fifth births at 27 and 28 months. Furthermore, it can be observed among women with intermediate and high school or college education in each cohort, that the birth intervals increase as birth order increases, but among women with at most primary education, the birth intervals stabilize after the third birth.4.3 Contraceptive Effects on Childspacing

Although the nature of the data on contraceptive use in this study is deficient, to examine fully the effect of contraception on fertility behaviour, it has been found in the previous chapter that ever-users show a slower pace of fertility than never-users after the third births. In this section, contraceptive effects on childspacing will be reexamined with women's age at time of survey being controlled. Specifically this section will examine whether or not women of different cohort and contraceptive status demonstrate the same pattern found earlier.

Table 4.4 gives the proportion of women having subsequent births at each successive duration and average

84

birth intervals at each birth order. As seen in the table, within each cohort, ever-users display a slower pace of fertility in the higher parities than never-users do. However, younger women who are ever-users contain a smaller proportion having subsequent births than never-users after second births while among older women, a smaller proportion of ever-users having subsequent birth than never-users became prominent only after the third births, as illustrated in Appendix - C Figure 6.

At a duration of 60 months, 85 per cent of younger women who had ever used contraception had third births and 77 per cent of them continued to have fourth births. Among younger cohorts who had never used contraception, 91 and 86 per cent of the women had third and fourth births within the same duration respectively. Among older cohorts, the proportion of ever-users having third and fourth births within 60 months are 89 and 81 per cent compared with never-users who have corresponding proportions of 88 and 85 per cent. As found earlier in this chapter, selectivity biases affect the results of this study particularly among younger cohorts at parity five. Probably for the same reason, younger women at parity five exhibit a higher proportion having subsequent birth than older women in each of the contraceptive use categories. Nonetheless, it is quite noticeable from the table that ever-users from younger cohorts have a slower pace of subsequent fertility than older cohorts. In contrast, among never-users, younger women tend to have a faster pace of having subsequent births than older women.

le 4.4

: Summary M

easures for

Birt

h Int

ervals

of Bir

th Order

2,3,4,5,

6 by

Contra

ceptive Use and

Age at the

Time

of the

Surv

ey, Philip

pines,

1983

nOs oj in »- s in co ojin s oo o d o d o

N.’ S-dcm

5oo

O'rooo'

Ko o

o£

N.O'*cm

cmco

IDTJl_O

.c

co

in

st

ro

'O r- rO O' O' oM -O COo d d

inincoo

CO rOd 00CM CM £

<4- v*- vO rO CM.— N- '— 'O ■IO -O CO CO O• ■ ■ • Mo o o o

indCM £N-

O O OoO'o

O'vdCM

fOin

m s sf ^ m '-sin in >o m «—m >o N co *- o- o• in cm ino o o o «-

-o o O' s. O' «sj S O sf ■ • -ocn >o co co o n . co• ■ • • CM CM <0o o o o

O' >o O O Vt o ">■co »■ m oo • • min s co co o s cm• ■ • • CM CM COO O O O «-

CM

in 'O3 fcd d §o

rsjf\joo

\n vt C\J C\J<\lro?M

00 CM 00 O' ro in in O'o o o

invdCM inCM

IDCO

Oomd

o

o3o

in-t

a»T3t_OJO*->

CO

in

st

ro

coinCMd

CMind

ininOo

ro

o

-oo roro

£o

'S3«£

3rod

inO'mo

inO''Oo o

vOK)

CMs

*- in -O in ro O

d dN.Ko

300d

ro3 cdCM

§

O' o *- «- m ro <-»CO 'O 'O CM • • 'Ocm in 'O N- co cm in• • • • ro ro «—o o o o «-

rs. t CM O O O' «- ro co ocm o K. r- ro o cm• ■ • • ro ro sto o o o «-

«- CM st O CM 00 ^ro m >o <- • • O'ro 'O N- co «— oo st. . . • ro cm 'Oo o o o T-

&22"O

£>!

st N- CM CM O'«— st in co • ■<t S co oo S in• • • • CM CMO O O O

CM

O'CMmo

o00d

oO'd

O'roO'doCM roCM

st st o ost ro cm vOin oo O' O' ro. . . . CM roCM 00

(/)o

(/>

o S (A<0 co aw 0(A(0S o S O£

° zCO b- X zE-4, '0 00 0 —CM fO >3* >0CQ CÜ CD CO H- iSE z

CDO)<roin

3inro

OJo

: Computed fro

m the

1983 Nat

ional

Demographic

86

The average birth intervals expressed in terms of the trimean further confirm that ever-users have a slower pace of fertility than never-users within each cohort in the higher parities. Among younger cohorts, ever-users have their third and fourth births at 3 0 and 3 5 months while never-users have them at 28 and 3 0 months respectively. Among older cohorts, however, ever-users have third and fourth births at 27 and 32 months while never-users have birth intervals of 28 months and 3 0 months for third and fourth birth intervals.

Evidence for ever-users of contraception to delay the subsequent birth at higher parities is more visible among younger cohorts than older cohorts. Fertility studies in the Philippines show that recent cohorts tend to prefer smaller family size (Cabigon, 1984). Possibly, younger women started using contraceptives much earlier during the childbearing period to space as well as to limit subsequent births, unlike older women who are likely to use contraception with the motive of limiting additional births. it should also be mentioned that younger women who are more educated tend to use the more modern methods of contraception more efficiently than older women.

Furthermore, it is expected that the impact of contraception will be seen more among women currently aged 15-3 4 than among women aged 3 5 and over at the time of the survey. This is because when the family planning programme services were launched in the country in the early 1970s, and became widely available, older women probably had already most of their children unlike younger

87

women who were in the early stage of family formation.Conversely, younger women who are never-users of

contraception show shorter birth intervals than their older counterparts. One possible cause of this pattern is the shortening of the duration of the post-partum amenorrhea among younger cohorts who tend to breastfeed for shorter durations than older cohorts. Data from the 1978 Republic of the Philippines Fertility Survey (RPFS) and the 1983 NDS show that older mothers in their mid-thirties or older tended to breastfeed their children two months longer than younger mothers (UPPI, 1985: 75) . In fact, postpartum amenorrhea in the Philippines declined from eight months in 1978 to a little over six months in 1983 (UPPI, 1985: 77).

4.4 Infant Mortality Effects on ChildspacinqThe tendency among women who had experienced infant

loss to have subsequent births much sooner than women who did not experience infant loss still persists within each cohort as illustrated in Appendix - C Figure 7. As shown in Table 4.5, among younger cohorts, 93 per cent of women whose first child survived during infancy have second births within 5 years and 76 per cent of women whose fourth child survived up to 12 months have fifth births in the same duration. In contrast, 98 and 87 per cent of women who experienced prior infant loss have second and fifth birth within 60 months. Similarly, among older cohorts, 93 per cent and 80 per cent of women who did not experience prior infant deaths have second and fifth births within 60 months while the corresponding proportions among women who

88

did experienced prior infant deaths are 95 and 84 per cent. The average birth intervals, as measured by the

trimean, give the same observation as found earlier within each cohort women who had experienced infant loss prior to the onset of the interval exhibit shorter birth intervals than their counterparts who did not experience prior infant deaths. Among younger cohorts, the birth intervals among women whose prior child survived the first year of life ranges from 25 months for the second birth to 35 months for the fifth births as compared with women whose prior child died during infancy of 18 and 26 months for the second and fifth births intervals respectively. This result gives further evidence that women experiencing infant loss particularly at lower parities tend to exertless or no effort to delay the subsequent births to replace their lost child quickly, resulting in shorter birthintervals. In fact, de Guzman's (1984) study found that among couples experiencing infant mortality at earlierparities, the attitude towards family planning in spacing or limiting subsequent births is less favourable. Couples who had experience infant mortality tended to delay the use of contraceptive and were less likely to use efficient methods of contraceptive than couples who did not have such experience. However, the result based on the tablepresented should be taken cautiously, because of the lack of control for important variables which are likely to affect the timing of subsequent births.

89

Comparing between cohorts, it can be seen that among women who did not experienced prior infant deaths, younger women show a slower pace of fertility than older women after the second births. Among women who did experience prior infant deaths, the reverse pattern is seen - younger women have a higher proportion having subsequent births in each successive durations and shorter intervals in each birth orders.

As mentioned earlier in this chapter, younger cohorts tend to exhibit a pattern of rapid subsequent births in early parities and have a tendency to limit additional births after the preferred number of children is obtained, which is lower in number than those preferred by older women. This tendency made childbearing of younger cohorts concentrated in a shorter period. Perhaps, because of this pattern, younger women who experience infant mortality have birth intervals shorter than those of older cohorts. Nevertheless, this pattern could also be a result of selectivity biases, which the present study failed to control. Younger women who had experienced infant mortality can be a select group in that they are more prone to have high risk of infant mortality at the same time as high fertility. If this is the case, younger women who experience infant mortality tend to reproduce more quickly than older women, not simply because they want to replace their lost child during infancy, but because they are morefertile.

le 4

.5

: Su

mm

ary

Mea

sure

s fo

r B

irth

In

terv

als

of

Bir

th O

rder

2,3

,4,5

,6 b

y S

urvi

vors

hip

of P

revi

ous

Live

Bir

th

in t

he F

irs

Yea

r of

L

ife a

nd A

ge a

t th

e Ti

me

of

the

Sur

vey,

P

hili

pp

ine

s,

1983

«-> 'S >.CO o

• i to > — 3<►— (/?

•O S- -J- CM vO IT» ^(M l/^ fO s j • ' CO<0 N- 00 O' f\J «— -vf

■ • • • CM CM wO O O O

N- vj- fO —>>

n >ö s co >o o rC• • • • CM CM w

cm oo -j- eg o vO in vQ N- 0 0

o o o o

in <- oo o eo vtin cm t— n . ■ «"O OO o o o o

• • • ■ CM CMO O O O

0 0 I O 00 (VJ > t< - O CO N-O CO CO o - o o

• ■ • • CM * -o o o o

O 00 m CM

o co m vO vj- vj-O «— to O' • •ro 'O N. s - ro o

• • • • ro roo o o o

-o cm in o >*■ oin co o S- • ■ro >o co co o. . . . CMo o o o

coCM

O O O O

o m Cw ^ (D

>o oo o • - *5rvj ro ^ >o *- oCO CD CO CO h- X

2

2N.

m vj- o> in in cm ^cm cm -o N- • • CM00 O' O O' S N r—

1 ' • I « - i - CMO O O O w

N- o m N. vn «—•— N. O IO • •co 'O s ooo o o o

00

o o co m o in în o .— 'O • • cmCM 'O N- in o o

• • • " CO CO -oo o o o «-

OroCOro

in vj- cm o o T - Ô' o> 0 « ro cm^ S CO O >» o• • • • CM CM O'

cm ro 'O oo t—00 *— NO O • • CMi n S S N ^ IM T -

• . . . CM CM t -O O O O

N 00 K1 -O (M r - -xO' «— «“ rO ' • CMx f S 00 CO <0 ' t CM

. . . . CM CM «—O O O O w

*— o 0> in N- roo cm s - og<0 oo oo oo *- «-. . . . CM CMo o o o

'— CM 00 O 'O O '-xcm ro o cm • • ' t<0 00 O' O' «-* O <0

. . . t CM CM «—O O O O w

O oo >0 O 1- CMco v* o m • • <gO OO O' O O' CO 'Oo o o o

OO N. 00 O' OO CMr*. 0 «- N-CM 1^ r . K x t r -

• ■ • • ro roo o o o

■•O O >*• x$- O * -O CM vl- Oro 'O N- co ro o

• • • • ro roo o o o

xfr "O N- >o o»CM in N- CMro >o r - oo o oo• • • • ro cmo o o o

>o «— ro «— >oo cm vj- egro N- co co co -o

• • • • CM CMo o o o

O' O O' N- ro >oN O' CO CMr co o in xj-

• • ■ ■ CM CMo o o o

~ c<3 CO Cw g CO

V* »o OO o ■— x>cm ro >1- o t_ <uoo co co co r— x

roCM

nI

roinro

8.22“S

0

>1

- s .

2o> u5

%<u*->

ro .84» T3

jQ 3 03t/>

C u)«/» — «or or0 °c <uin01 <u0) JZ

w * 0)

OG\

91

CHAPTER 5SUMMARY AND CONCLUSION

Fertility behaviour is a complicated process with numerous factors needed to fully understand its process. Urban-rural fertility differentials exist because of the differences in the socio-economic structure and the rapid modernization of the urban areas affecting the couples' decision of ideal family size and spacing of births. The Philippine experience a declining fertility since the early seventies, concurrent to changing fertility behaviour among filipino women.

This study examined fertility differentials and changes in terms of childspacing practices among Filipino women of different socio-economic characteristics. A life table technique was utilized to calculate the proportion of women having a subsequent birth and the length of the birth intervals, and these measures were then used to analyse fertility behaviour. In particular, the analysis considered the intervals after the first birth up to the intervals after the fifth birth among women who are currently married and married only once. The factors considered in the analysis were current place of residence, education, contraceptive use, experience of infant mortality and women's age.

92

5.1 Summary of FindingsChildspacing differentials among women currently

residing in Metro Manila, other urban areas and rural areas were examined in Chapter 3. A smaller proportion of women having further births and longer average birth intervals among women in Metro Manila and in other urban areas were found, as compared with women in rural areas. This further support previous studies that urban fertility is lower than rural fertility.

The pattern of childspacing among rural women is more consistent from the second to the sixth birth than among Metro Manila or other urban women. The proportion having further births among rural women, by parities, decreases only gradually. By contrast, Metro Manila and other urban women have faster pace of fertility at lower parities then exhibit substantial decline from the third birth onwards. Rapid decline in the proportion of women having further births after the second birth is observed more among Metro Manila women. The findings suggest that differences in childspacing patterns are likely to occur among populations of different socio-economic structure and level of urbanization. It appears that in Metro Manila and other urban areas fertility follows the modern pattern of reproduction evident among developed countries and some of the developing countries.

Compared with rural areas, a smaller proportion of women in Metro Manila and other urban areas continue to have subsequent births after the third births. This indicates a smaller family size is preferred among women

93

dwelling in the urban areas than rural areas. On the other hand, rural women exhibit a consistent higher proportion having subsequent births up to sixth birth, suggesting that relatively more children are desired by rural families, which are likely to be engaged with agricultural activities. Moreover, though Metro Manila and other urban women tends to have shorter second birth intervals, generally women in urban areas have longer average birth intervals at higher parities than their rural counterparts.

An examination of the effect of education, contraceptive use, and infant mortality among women of different types of place of residence was also made. The study confirms the inverse relationship of education and fertility in all areas, but it was more pronounced in Metro Manila. Women with higher education have lower fertility, as indicated by the smaller proportion having subsequent births at all parities and longer birth intervals after the second birth.

The effect of experience of contraception on the quantum and timing of fertility becomes noticeable only at higher parities. At lower parities, different patterns were found in each area. In other urban and rural areas, the proportion of ever-users having further births is higher than never-users only at the second birth. In Metro Manila, the proportion of ever-users having further births were found to be smaller than never-users only after the third birth. The average birth intervals show the same pattern - longer birth intervals among ever-users than never-users at higher parities.

94

The study also provides further empirical evidence of the effect of infant mortality on fertility. Generally, women who did not experience prior infant death have a smaller proportion having subsequent birth, at all parities, and longer average birth intervals than women who did experience prior infant death. The small number of cases in Metro Manila and other urban areas whose previous child died during infancy made comparison with rural areas not possible.

The effect of cohort differences on the pace of subsequent fertility was examined in Chapter 4. Younger cohorts, as defined in this study, consist of women 15-34 years of age at the time of the survey while older cohorts consist of women aged 35 and over. Younger cohorts were found to have lower fertility than older cohorts as indicated by a smaller proportion having subsequent births than older women at the third to fifth births. An inconsistency observed at the sixth birth among the younger cohort is likely to be a spurious result of the selectivity. On the other hand, the observed pattern of a second birth sooner after the first birth among younger women is not merely a selectivity bias but rather suggests a recent change in fertility behaviour among younger cohorts.

Childspacing differentials between the younger and older cohorts were examined further considering factors such as education, contraceptive use and experience of infant mortality. It was found that in general women in the younger cohort with at least intermediate education

95

have a smaller proportion having subsequent births at higher birth orders than older women of the same educational level. No differences nor any general pattern are observed among women with at most primary education in the younger and older cohorts. Nonetheless, the inverse relationship of education and the pace of fertility at higher parities remain within each cohort. A similar pattern is revealed by the average birth intervals.

Effects of the use of contraception are observed in each cohort, though it is more visible among younger cohorts. Within each cohort, except at the second birth for younger cohorts and up to the third birth for older cohorts, ever-users have a lower proportion having subsequent births and longer birth intervals than never- users. The use of contraception to delay the subsequent births at higher parities is more noticeable among younger cohorts than older cohorts. This pattern is consistent with previous studies showing that recent cohorts prefer a smaller size of family* and tend to use contraception much earlier during the childbearing period, to space as well as to limit the coming of additional births. Older cohorts use contraception more for the motive of limiting subsequent births after the preferred number of children is reached.

Within each cohort the effect of infant mortality on fertility, was that for women who had experienced infant death, the proportion having subsequent birth was higher than of women who did not experience the loss of a child during infancy. Furthermore, infant death leads to

96

shortening of the average birth interval. This pattern is clearly seen at earlier parities, where couples tend to replace their lost child more quickly. Comparing the cohorts, it was found that after the second birth, among women who did not experienced prior infant death, younger women show a slower pace of fertility than older women, but among women who did not experienced prior infant loss, the reverse pattern is seen.

To summarize, the main findings of the present study are as follows:

- The proportion of women having a subsequent birth by each duration declines with parity in Metro Manila, other urban areas and rural areas. In Metro Manila and other urban areas, the proportion of women having a subsequent birth tended to decline faster after the second birth. In rural areas, the proportion of women having subsequent births from parity 1 to parity 5 shows only a gradual decrease. In all areas, the average birth interval measured by the trimean increased with birth order up to the fourth birth then stabilized. Compared with rural women, women in Metro Manila and other urban areas have longer birth intervals after the second birth.

- Women with more education exhibit lower proportions having subsequent birth and lower birth intervals at all birth orders (except at the second birth) and in all areas. The negative effect of education on the rate of childbearing is particularly strong at higher birth orders. Even after controlling for educational level, rural women still exhibit a higher proportion having a subsequent birth

97

at higher parities. When women's age is controlled, the inverse relationship of education and fertility within each cohort persists. However, the effects of higher education on the pace of subsequent birth and birth intervals was more visible among younger cohorts than older cohorts. Younger women tended to have smaller proportions having subsequent births and longer birth intervals at higher parities than older women.

- In other urban and rural areas, the proportion of ever-users of contraception having further births is higher than never-users only at the second birth. In Metro Manila, however, the proportion having subsequent births among those who had used contraception is higher than those women who had never used contraception at both the second and third births. The average birth intervals show the same pattern. There is indication that the effect of experience of contraceptive practice on the quantum and timing of fertility operates more at the higher parities. The effect of contraceptive use on the rate of childbearing remains even when women's age is controlled. However, among ever-users, younger women exhibit a lower proportion having subsequent births than older women.

- For each birth order the death of the previous child within the first year of life tended to increase the proportion of women having subsequent birth, and shortened the gap between births. Information from rural areas further confirms the effect of infant mortality on fertility behaviour. When women's age is controlled the same effect of infant mortality on birth intervals is

98

observed. Younger women whose prior children survived infancy have a lower proportion having subsequent births, after the second child, and higher average birth intervals from the third to higher births, than older women.

5.2 Concluding RemarksThe study gives empirical evidence of the childspacing

differentials and changing fertility behaviour among Filipino women. The results suggest that factors related to modernization and the urban way of life influencing childspacing practices among Metro Manila and other urban women cause the differences between Metro Manila, other urban, and rural fertility. Metro Manila and other urban women seem to have fewer children sooner, tending to shorten their period of active childbearing, whereas a high proportion of rural women continue to have subsequent births. The smaller proportion of younger cohorts progressing from third to fifth births within five years and a general tendency to have longer birth intervals than older cohorts suggests the country is experiencing a recent change in the family building process.

The findings concerning education and fertility conform with the results of other studies conducted elsewhere. Although apparent only after the second birth, an inverse relationship between education and the rate of childbearing exists in each area. Further analysis also shows that the impact of education on fertility behaviour is more noticeable among younger cohorts than older

99

cohorts. This pattern is to be expected as younger women enjoyed better opportunities to profit from their education in taking advantage of family planning services available, not simply to limit births but also to space their births.

The effect of contraceptive use is seen to operate only at higher birth orders. This is a substantial indication that most Filipino women tend to use contraception not so much to space births but to limit their births after their family has become disagreeably large. This might be a concern for the national family planning programme in the country. To be more effective, their programme would need to concentrate on efforts in promoting the use of contraception both for spacing and limiting further births. Availability of effective methods of contraception among women who want to have few children and want to prevent exessive births is essential. On the other hand, women who want more children should be helped to space their children with longer intervals and advised about the importance of well spaced births. The programme should also exert efforts to prevent discontinuation of contraceptive use among women at risk of childbearing. Accessibility, affordability and promotion of efficient use of modern methods of contraception, particularly in the remote rural parts of the country, might also be a concern for program implementors. This could probably lessen, to some extent, the marked differences in the proportion of ever-users having further births observed between women inthe urban and rural areas.

100

The study also noted that women at any parity tend to exhibit higher subsequent fertility and consequently shorter birth intervals when experience infant mortality. This pattern persists even after controls for age were introduced. Perhaps the presence of infant death has " a direct effect on fertility either because mothers tend to replace children who have died, or because women whose children who have died have reduced period of breastfeeding and amenorrhea, resulting in shorter intervals" (Rodriguez and Hobcraft, 1980: 34). Furthermore, evidence of theeffect of infant mortality on fertility is observed to be quite strong at earlier parities. At earlier parities where the need to replace a deceased child is more urgent, women who experienced infant death probably delayed their first use of contraception and were unlikely to use the efficient methods. This implication, although notconclusive, supports the "child-survival hypothesis" increased child survivorship could contribute to the decline in fertility through increased motivation towards the use of family planning.

Further research in several areas would be necessary to understand more the pattern and causes of childspacing differentials among Filipino women. For instance, women's ages at the start of childbearing and the elapsed time before the first birth, which were not considered in the present study, should be examined in relation to subsequent fertility in future research. These two factors - age at entry to marital union and the timing of first birth could give additional information on the pattern of childbearing

101

and completed family size. Further research is also suggested, taking into consideration other variables such as foetal mortality and women's work status, which could play an important role in childspacing practices.

102

R E F E R E N C E SARTHUR, W. B. and G. McNICHOLL

1978 'An Analytical Survey of Population andDevelopment in Bangladesh7in Population and Development Review. Vol.4, no.l (March), pp 23- 80.

BALAKRISHMAN, T.R.1978 'Effects of Child Mortality on Subsequent

Fertility of Women in Some Rural and Semi-Urban Areas of Certain Latin American Countries7 in Population Studies Vol. 32 no.l, pp. 135-145.

BOGUE, D.J. and E.J. BOGUE1980 Comparative Birth Interval Analysis in Family

Planning Research and Evaluation Manual no. 15.CABIGON, J.V.

1984 'Determinants of the Timing of First Births in the Philippines7in Fertility in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978. ed. by L.T. Engracia,C.M. Raymundo and J.B. Casterline,pp. 101-116, Voorburg: International Statistical Institute.

1984a 'Factors affecting use and non-use ofcontraception7in Fertility in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978. ed. by L.T. Engracia,C.M. Raymundo and J.B. Casterline,pp. 149-172, Voorburg: International Statistical Institute.

1984b 'The Timing and Spacing of Birth: the Philippines case7in Fertility in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978. ed. by L.T. Engracia,C.M. Raymundo and J.B. Casterline, pp. 61-73, Voorburg: International Statistical Institute.

1985 'Fertility and Contracption in the Regions7 in Philippines Population Journal Vol. 1 no. 4 (December), pp. 30-66.

n.d. Preliminary Report of the 1983 NationalDemographic Survey CNDS)in the Philippines, University of the Philippines Population Institute (UPPI).

103

CABIGON, J.V and HUFANA1984 'Determinants of Age at First Birth' in Fertility

in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978. ed. by L.T. Engracia, C.M. Raymundo and J.B. Casterline,pp. 101-116, Voorburg: International Statistical Institute.

CALDWELL, J.C.1980 'Mass Education as a Determinant of Timing of

Fertility Decline' in Population and Development Review. Vol.6, no.2 (June), pp.225-256.

CALDWELL, P. AND J.C. CALDWELL1981 'The Function of Child-Spacing in Traditional

Societies and the Direction of Change' in Child-Spacing in Tropical Africa: Traditions and Changes,ed. by H.J. Page and R.Lesthaeghe, pp. 73-92, Belgium: Academic Press.

CHANDER, R., V.T. PALAN, N. AZIZ and TAN BOON AUN1977 The Malaysian Fertility and Family Survey 1974

The First Country Report. Government Printers, Malaysia, Kuala Lumpur.

CHOWDHURY, A.K.M.A., A.R. KHAN and L.C. CHEN1976 'The Effect of Child Mortality Experience on

Subsequent Fertility: in Pakistan and Bangladesh' in Population Studies. Vol.30, no.2, pp. 249-261.

CLELAND, J. and S. RUTSTEIN1986 'Contraception and Birthspacing' in International

Family Planning Perspective. Vol.12, no.3 (September), pp. 80-82.

CONCEPCION, M.B.1978 'Population Policy and Family Planning' in

Country Mongraph Series No. 5 Population of the Philippines. pp. 172-190, Bangkok: Economic and Social Commission for Asia and the Pacific.

DAVIS, K. and J. BLAKE1956 'Social Structure and fertility: An Analytical

Framework' in Economic Development and Cultural Change. Vol. 4, pp. 211-235.

DONALDSON, P.T. and D.J. NICHOLS1978 'The Changing Tempo of Fertility in Korea' in

Population Studies. Vol. 32, no.2 (July), pp. 231-249.

104

ENGRACIA, L.T., M.M. MÖRTEL and L.B. NARTATEZ1984 'Accessibility of family planning, community

develepment and their impact on birth rate of rural Philippines'in Fertility in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978. ed. by L.T. Engracia, C.M. Raymundo and J.B. Casterline, pp. 175-189, Voorburg: International Statistical Institute.

ENGRACIA, L.T., C. RAYMUNDO and J.B. CASTERLINE (ed)1984a Fertility in the Philippines Further Analysis of

the Republic of the Philippines Fertility Survey 1978. Voorburg: International Statistical Institute.

ENGRACIA, L.T. and Y. XIM1979 'Fertility Differentials Among Filipino Women' in

UNFPA-NCSO Population Research Project Monograpgh. no.15, Manila:NCSO

FARID, S.M.1974 'On the Tempo of Childbearing in England and

Wales' in Population Studies. Vol. 28, no.l, (March), pp. 69-84.

FAWCETT, J.T. and S.E. KHOO1980 'Singapore: Rapid Fertility Transition in a

Compact Society' in Population and Development Review. Vol.6, no.4 (December), pp. 549-580.

FEENEY, G.1983 'Population Dynamics Based on Birth Intervals and

Parity Progression' in Population Studies Vol. 37 no. 1, pp. 75-89.

FERRY, B. and D.P. SMITH1983 'Breastfeeding Differentials' in WFS Comparative

Studies no. 23 (May).FREEDMAN, R. and J. CASTERLINE

'Nuptiality and Fertility in Taiwan' Paper presented at the Seminar on Nuptiality and Its Effect on Fertility, IUSSP Bruges, Belgium, January.

GO, E.M.1983 Breast-Feeding in the Philippines: Level and

Differentials. Canberra:unpublished Master thesis.

GUILKEY, D.K., R.R. RINDFUSS, B.M. POPKIN, J.S. AKIN and V. PAQUEO

1988 'Childspacing in the Philippines: the Effect ofCurrent Characteristics and Rural Development' in Population Studies Vol. 42, pp. 259-273.

105

DE GUZMAN,E.A.1984 'The Effect of Infant Mortality on Fertility in

the Philippines'in Fertility in the Philippines Further Analysis of the Republic of the Philippines Fertility Survey 1978, ed. by L.T. Engracia, C.M. Raymundo and J.B. Casterline, pp. 123-130, Voorburg: International Statistical Institute.

1978 'Trends and Differentials in Fertility'in ESCAP Country Monograph Series NO. 5 Population of the Philippines. Bangkok:UN.

HOBCRAFT, J. and J. MCDONALD1984 'Birth Intervals' in WFS Compartive Studies no.28

(March).KANDIAH, V.J.

1981 'Childspacing in Peninsular Malaysia: 1966/67 and 1974. Michigan: unpublished PhD. thesis.

KLINGER, A.1977 'Fertility and family Planning in Hungary' in

Studies in Family Planning. Vol. 8., pp. 166-176.KNÖDEL, J.

1978 'European Population in the Past; Family Level Relations' in The Effects of Infant and Child Mortality on Fertility. New York: Academic Press.

LESTHAEGHE, R., H.J. PAGE and O. ADEGBOLA1981 'Child-spacing and Fetility in Lagos' in

Child-Spacing in Tropical Africa:Traditions and Changes.ed. by H.J. Page and R.Lesthaeghe, pp. 147-180, Belgium: Academic Press.

LONDON, K.A.1985 'Fertility and Family Planning Surveys: An

Update' in Population Bulletin no.8.LUCAS, D.

1980 'Fertility'in Beginning Population Studies, ed.by D. Lucas, P. McDonald, E. Young and C. Young, pp. 63-92, Canberra: The Australian National University.

MERRICK, T.M. and PRB STAFF1986 'Fertility Decline in LDCs' Population Bulletin,

Vol.41, no.2 (April), pp. 30-38.MOSLEY, W.H., L.H. WERNER, and S. BECKER

1982 'The Dynamics of Birth Spacing and MaritalFertility in Kenya' in WFS Scientific Reports no. 30 (August)

106

NATIONAL ECONOMICS AND DEVELOPMENT AUTHORITY (NEDA)1983 'An Introductory Guide to Population and

Development Planning. Pasig: NEDA.NATIONAL CENSUS AND STATISTICS OFFICE (NCSO)

1984 'Statistical Handbook of the Philippines 1984.Manila: NCSO.

NATIONAL CENSUS AND STATISTICS OFFICE (NCSO)1983 1980 Census of Population and Housing Philippines

Vol. 2 Natinal Summary. Manila: NCSO.NATIONAL CENSUS AND STATISTICS OFFICE (NCSO), UNIVERSITY OF THE PHILIPPINES POPULATION INSTITUTE (UPPI), POPULATION COMMISSION (POPCOM), and NATIONAL ECONOMIC AND DEVELOPMENT AUTHORITY (NEDA)

1979 '1978 Republic of the Philippines Fertility Survey - World Fertility Survey CRPFS- WFS).First Report. Manila: NCSO, UPPI, POPCOM, NEDA.

NATIONAL ECONOMIC AND DEVELOPMENT AUTHORITY (NEDA)1984 Updated Philippine Development Plan 1984-1987.

Manila.PEBLEY, A.R. and S. MILLMAN

1986 'Birthspacing and Child Survival' in International Family Planning Perspectives. Vol. 12, no.3 (September), pp. 71-79.

PERALTA, A.R. and M.C. LIGAN1975 Philippine Population: Implication. Programs and

Politics. Manila:University of East Press.PRESTON, S.H.

1978 'Introduction'in The Effect of Infant and ChildMortality on Fertility, ed.by. Preston, New York: Academic Press.

REPUBLIC OF THE PHILIPPINES (RP)1987 Medium - Term Philippines Development Plan 1987 -

1992. Manila.RINDFUSS, R.R, L.L. BUMPASS, J.A. PALMORE, M.B. CONCEPCION, D.N.L. Bt DATO' ABU BAKAR, S. GAMAGE, S. SAENGTIENCHAI and N.I. KIM

1984 'Childspacing in Asia: Similarities and Differences' in WFS Compartive Studies. No. 29 (August).

RODRIQUEZ, G. and J. HOBCRAFT1980 'Illustrative Analysis: Life Table Analysis of

Birth Intervals in Colombia' in Scientific Reports. no.16 (May).

SAENGTIENCHAI, C.1985 Childspacinq Practices among Thai Women.

Canberra:unpublished Master Thesis.

107

SANTOW, M.G.1978 'A Microsimulation of Yoruba Fetility' in

Mathematical Bioscience. Vol.42, pp.93-117.SRINIVASAN/ K.

1980 'Birth Interval Analysis in Fertility Survey7 in WFS Scientific Reports, no.7 (February).

SRINIVASAN, K., P.H. REDDY and K.N.M. RAJU1978 'From One Generation to the Next: Changes in

Fertility, Family Size Preferences and Family Planning in an Indian State between 1951 and 19757in Studies in Family Planning. Vol. 9, pp. 258 - 271.

TAYLOR, C.E. and M.F. HALL1967 'Health, Population, and Economic Development7 in

Science, no. 157, pp. 651-665.TAYLOR, C.E•, J.S. NEWMAN and N.Ü. KELLY

1976 'The Child Survival Hypothesis7 in Population Studies, Vol.30, no.2, pp. 263-278.

TRUSSEL, J., L.G. MARTIN, R. FELDMAN, J.A. PALMORE,M. CONCEPCION and D.N.L. Bt DATU' ABU BAKAR

1985 'Determinants of Birth-Interval Length in thePhilippines, Malaysia, and Indonesia: A hazard- Model Analysis7 in Demography. Vol.22, no.2 (May) pp. 145-168.

THORTON,1983

A. and D. FREEDMAN'The Chnaging American Family7 in Population Bulletin. Vol.30. no.4 (October).

UNITED NATIONS (UN)1979 'Republic of the Philippines Fertility Survey,

1978 A summary of Findings7in World Fertility Survey, no.15 (November).

UNIVERSITY OF THE PHILIPPINES POPULATION INSTITUTE (UPPI)1983 Philippines Population Data: An Update. Quezon

City: UPPIVAN GINNEKEN, J.K

1978 'The Impact of Prolonged Breast-feeding on Birth Intervals and Postparthum Amenorrhea7 in Nutrition and Human Reproduction, ed. by. W. Henry Mosley , pp. 179-195. New York and London: Plenum Press.

WHELPTON,1964

P.K.'Trends and Differentials in Spacing of Birth7 in Demography. Vol.l, no.l, pp. 83-93.

YOUNG, C.1977 'Spacing of Children and Changing Patterns of

Childbearing7in Journal of Biosocial Science. Vol.9, pp. 201-226.

108

APPENDIX - A

The concept used by the 198 3 National Demographic Survey (NDS) in classifying areas as urban is based on the 1980 Census of Population. According to the concept used in the 1980 Census, which also was based on urban-rural definition of the 1970 and 1975 Censuses, urbanized areas consist of ( NCSO, 1983:xi):

1. In their entirety, all cities and municipalities having a population density of at least 1,000 persons per square kilometer.

2. Poblaciones or central districts of municipalities and cities which have a population density of at least 500 persons per kilometer.

3. Poblaciones or central districts (not included in 1 and 2), regardless of the population size, which have the following:a. Street pattern, i.e., network of streets in

either parallel or right-angle orientation;b. At least six establishments (commercial,

manufacturing, recreational and/or personal services); and

c. At least three of the following:1) A town hall, church or chapel with religious

services at least once a month;2) A public plaza, park or cemetery;3) A market place or building where trading

activities are carried on at least once a week;

4) A public building like a school, hospital, puericulture and health center or library.

4. Barangays having at least 1,000 inhabitants which meet the conditions set forth in 3 above, and where the occupation of the inhabitants is predominantly nonfarming or fishing.

In areas not falling under any of the above classifications are considered rural.

The

Cumu

lati

ve P

ropo

rtio

n of W

omen

of

Birt

h Or

der

2,3,

4,5,

6 wh

o ha

ve S

ubse

quen

t Bi

rth

by B

irth

Int

erval, E

ducation!

and

Plac

e of R

esid

ence

, Ph

ilip

pine

s, 1983 (

in p

erce

nt)

«- N- «- »-'OCÜ -sT ovfed'O -d d d n! <»Kl AJ AJ ' O ' O ' Ola vr oCO CO N

• - M S eo O' «- m oo •-CO OO in T- K1 -O M -O COM (M (\J 0 LA CO N 'O

O IO sfS Kl sf CO CO s

OO AJ O' >f M >0 M 'O M « ♦ « - » -• t ■ * * I I ■ • * N.n^ « - «— o o o k i «— «— eg s eo ^« Ki m N.'Oin co co s co co s c>

CO CM S «— 00 N- LA O' LA N- O -»• '-no n ! -o m co s -o ~o oo•4- Kl Kl N - S . nO 00 CO N-

o o «- *- n- *- •— N- O'O O' O' O' Kl O' O' AJ 00m \f sf s co s co o oo

5 5 SS §>4-w

N» ro coin in mO' O' O LALA

M 00 M O' CM CO Nf in >- N- -4. . . . . . . . . . . . AJLA NO CM NO LA LA N- O Kl O nO nO O'rn CM M -O LA LA N - N - - 0 CO N -O nO

Kl LA O CO CO N O LA «-> d do n - * — o oo in «— «— ki >4 N»CM CM m NO LA LA CO S O CO N vO

NO «— NOr d

la no aj o Nt in o vf oN Al O S 'J IA O' O' 4in in m o o ia n- -o

-4 O' 00 C O ' O N - ki «- o<! o oo d rd d »g nJ n(O'l-to K- no co co N

-4 aj o>4 Kl N*00 00 -0

■4 oo OO O' AJO' 00 oo <XD

<T>ocH

*- O LA I O O - 1 - r- CO AJla O' oo o o 4 o co la4 4 4 to N N 00 a 00

AJ 'O nOKl AJ O' O O' oo

«£

00 O0 N- 00 O LA CM in O AJ 4 *- ô ki 4 d co »— ro O' n ! d d 4 oin in aj -j in si -o -o 4 -o n - la ki

IO O' O' O >— oo N - O O O nO N - *-*~0 aj d n ! d «- cm in co f o o - o 'Sm in aj in in nJ N - nO la oo N d -4

— « - o o o o in in la o n- ki o n - * — ^C LA CO n— O CO r- 4 4 O' AJ O' AJ O'co m in in n- ' O la oo n la o >o -o

0 ^ 0 ( M O O N . 4 AJ AJ i a N-'O £in aj 4 oo in o 4 d *- d ed d ki4 LA Kl N . S . ' O 00 00 N- CO CO N O'

8.*2

'S

I

>I

N - O O O O t- Kl O O' O' AJ Kl Kl £O' 00 Kl nO AJ LA CO nQ LA LA O O' Kl4 LA LA N - O O N . 00 00 CO O O' CO «—

-C CO■M >•t *CD <M

>* OL_ 4-#<3 <0 <L>

i l lV« i- o tl Cl u C *-> ' x O C Vj

>> oL. 4-» >■ <uc_ *-> >■ <uC— J->§ <l>O) g CO <uo> g CO <vO)

1 1 1 1 —1 1

01 <ö ou oT <ü "o 4> «3 oc *->o c cn c *->

o c c *-> o c c/>c •- JO c •- c — □C

dKl dÜ

v>OJ(ACOo

: Co

mput

ed f

rom

the

1983 N

atio

nal

Demo

grap

hic

The

Cum

ulat

ive

Pro

po

rtio

n o

f W

omen

of

Bir

th O

rder

2,3

,4,5

and

6 w

ho h

ave

Subs

eque

nt

Bir

th b

y B

irtl

and

Pla

ce o

f R

esid

ence

, P

hili

pp

ines

, 19

83

(in

per

cen

t)

o(/>Z>

u2•M

o

CM

01

■o 28

.3

31

.7

59

.9

67

.7

70

.3

81

.0

76

.8

86.6

(23

23)

CM fO cm ro ro cm ■vT 0O

0) in coCM

rorO CM N. 'O o>*• o n. r^ o- in

N . CO

"O ro

o^1- R

ural

Klro 33

.4

65

.9

68.6

76

.9

82

.3

82

.9

87

.6 38:-C

4-*

ro

L.N- CM ■vf in -J- N. -o ro

COrO vdCO C>ro id fd ro in co co cö o c5 o i

vT

'T -o o o S - O in in ,-s

(M rOin 3 co co" 00 N. CM O O CO

>o CM O O

lO>oininw

CO vO -o o 00 o CM O

Os 3

in cm in o i fd ^ n! r«. n. sw

ro ro in co in inL.0) in <>

CMC>CM

<>in in i pd CM s !r^ is.CMCMO

"Ü -O wL.

3

o U

o 31

.8

33

.2

59

.7

63

.0

71

.4

75

.6

76.1

79

.8

S t-C

•M

C.

COrO

CM

35

.7

69

.7

69

.8

80

.6

81

.9

86.1

87.1

(17

79)

(M

52

.3

43

.0

80

.0

72

.4

89

.8

82

.7

93

.5

87

.6

(21

17

)

in CM O CM O in in cm<5

5CMCM

O in vr -r in cm in - o s id ro

o in o in co ro o o /•>

c .in

COrs:CM CM

o in in vj- CM *— O -O

o ^-O N-in>ovT

0)

T3 m

c- * “ V * o in O 00 00 « - /-N

O vj-0

5inrO in in - o

in o ■O Is- O CM N - C -

X-C

•M

L. rO

41

.0

38

.2

65

.8

65

.4

76

.8

75

.6

81

.3

79

.6

C93

5)

CO

>o o ro « - in O N-CM

inin SCO vj-r - n . s s

cmO CO

ro

L. •

9)> «

U U(0

T J "S > > > (0

u

5□L.

l_

CO

DM CM

-Öro CO3 z

8.■22_g

I

>I

orH«H

: Co

mpu

ted

from

the

198

3 N

atio

nal

D

emog

raph

ic

Cum

ulat

ive

Pro

por

tion

of

Wom

en o

f B

irth

Ord

er 2

,3,4

,5,6

who

hav

e Su

bseq

uent

Bir

th

by B

irth

In

terv

al,

S

urv

ivor

ship

of

the

Pre

vivo

us L

ive

Bir

tl

he

Fir

st

Yea

r of

L

ife

and

Pla

ce o

f R

esid

ence

, P

hili

pp

ines

, 19

83

(in

per

cen

t)

i n - o 00 so fO - o i n r o2

>oco - i CM VO

f n cd SO N -

sO i n (V- CO

c m eg c o 2 FO

CMw

CO s . i n i n s o o O N -Lm

i n0> fOcm i n 3 £

sO v i S - CO

rv j mGO CO

Oa> r o

L.

ov i CO 3

1.9

63

.5

66

.5

83

.9

79

.2

91

.5

85

.1

93

.4

8OfO

«C1

Lm CO o o o fO -O m m

CDm

O CM fO -O

CM vT N . CO 2S o i

O CM CO Os

00

00

w

CM CM fs . K l i n co CM 0

CM00 OO NT S -

o ’ Os co co

O f * l O O '

v i s d O ' O '

m0i ni n

CM O ' i n i n s i i n 0 c o

-Oo o ’CM O

rsJ c m i n fs-

O ' CM sO fv . S R §

w

in s. i n co * - SO s— c o /■■X

t -i n

eo v rCM v i

fs~ i n i n >o

0 eg rv . <5

v i c ors~ c o

0

00)

wTJ

Lm §o

-O CO cn c o * - «— i n sO fs -

v i3 o 'O

fO -OO ' v fi n co

CM O fs . O '

sO i n fs . 0

O

JZ w

4-»Lm • o o T - m i n v i c o «— £

fon ! O ' fO - o

O ' CMsO co

ed 1—00 O '

i n c d co 0

CDw

o i n eo k * 00 v i- O ' fs . x-s

CMCO sdVi- SO

sO m(S. CO

- d i dCO Os

c d n ! O O '

00

i

eo m * - v i- v i- i n co sO

oc o n !CM CM

C> sov i- fO

c o i nm v i

fO m SO sO

0

w

s o fO Is- m CO fO 0 CM ✓“N

Lmi n

- d - dCM Vi-

od s dv i- «4-

CM s O - O v i 8 R

fO0

0) <0 *“OL.

o '-#■

sXo

33

.7

45

.5

58

.4

65

.9

66

.5

72

.7

70

.7

72

.7

(69

5)

-C1

4-»

«_m 4

0.0

46

.2

65

.7

67

.7

76

.6

74

.2

81

.2

80

.6

(93

7)

CO

53.

1

80

.8

76

.3

96

.2

85

.9

96

.2

89

.8

96

.2 5r o

fMw

> > > >

a ‘ 5 >

i oCO

c_

co

L

CO

u

10

o «5 > *->

0) 4->> o

01> o

0 *-> > 0

0) *-> > 0

C/>

L.’ > c u -O

— c' > c t- T3

— cC/>a>

3 — 3 3 — 3 eg

1/5 ° CO Q M Q Q U

x : >

Ü r 5sOr o

CO3 z

aJ

8.22■§

£>I

rHrH

: Co

mpu

ted

from

the

19

83 N

atio

nal

D

emog

raph

ic

APPENDIX C 112

FIGURE 1 : LIFE TABLES BIRTH FUNCTIONS BY PARITIES AND PLACE OF RESIDENCE

METRO MANILA

MONTHS SINCE PREVIOUS BIRTH

OTHER URBAN


RURAL


APPENDIX C 113

FIGURE 2 : LIFE TABLES BY BIRTH ORDER, EDUCATIONAL LEVELAND PLACE OF RESIDENCE

METRO MANILA OTHER URBAN RURAL

NONE/PRIMARY INTERMEDIATE ---- HS or COLLEGE

APPENDIX114- e

FIGURE 3 : LIFE TABLES BY BIRTH ORDER, CONTRACEPTIVE USEAND PLACE OF RESIDENCE


EVER USED NEVER USED

115APPENDIX - C

FIGURE 4 : LIFE TABLES BY BIRTH ORDER, SURVIVORSHIP OF PREVIOUS LIVE BIRTH AND PLACE OF RESIDENCE


FOURTH BIRTHFOURTH BIRTH FOURTH BIRTH

FIFTH BIRTHFIFTH BIRTH

SURVIVE DID NOT SURVIVE

APPENDIX - C 116FIGURE 5 : LIFE TABLES BY BIRTH ORDER, EDUCATIONAL LEVEL

AND AGE AT THE TIME OF THE SURVEY

FOURTH BIRTHFOURTH BIRTH

FIFTH BIRTH

SIXTH BIRTHSIXTH BIRTH

NONE/PRIMARY .. INTERMEDIATE --- HSOR COLLEGE

APPENDIX - CFIGURE : LIFE TABLES BY BIRTH ORDER, CONTRACEPTIVE USE

AND AGE AT THE TIME OF THE SURVEYAGES 1 5 - 3 4 AGES 35 - 49

SECOND BIRTH SECOND BIRTH

THIRD BIRTHTHIRD BIRTH

FIFTH BIRTH FIFTH BIRTH

SIXTH BIRTHSIXTH BIRTH

PVER USED NEVER USED

APPENDIX - C 118FIGURE 7 : LIFE TABLES BY BIRTH ORDER, SURVIVORSHIP OF

PREVIOUS LIVE BIRTH AND AGE AT THE TIME OF THE SURVEY

AGES 1 5 - 3 4 AGES 35 ~ 4V

THIRD BIRTHTHIRD BIRTH

FOURTH BIRTHFOURTH BIRTH

Documents

CHILDSPACING OF CURRENTLY MARRIED FILIPINO WOMEN MARRIED ... · CHILDSPACING OF CURRENTLY MARRIED FILIPINO WOMEN MARRIED ONLY ONCE by Jesusa T. Gascon A thesis submitted in partial