Trends in Breast Feeding Among American Mothers PUBS/A21.pdf · 2018. 10. 6. · Datafromthe NATIONALSURVEYOFFAMILYGROWTH Series 23 Number3 Trends in Breast Feeding Among American

Data from theNATIONAL SURVEY OF FAMILY GROWTH

Series 23Number 3

Trends in Breast FeedingAmong American Mothers

Statistics based on data collected in 1973 are presented on breastfeeding of first- and second-born babies. The percentages of moth-ers who breast fed and who breast fed for 3 months or more aredistributed by year of the mother’s birth and year of the baby’sbirth. These distributions are shown separately for women classifiedby race or ethnicity, geographic region, and various socioeconomicvariables.

DHEW Publication No. (PHS) 79-1979

DEPARTMENT OF HEALTH, EDUCATION, AND WELFAREPublic Health Service

Office of Health Research, Statistics, and TechnologyNational Center for Health StatisticsHyattsville, Md. November 1979

NATIONAL CENTER FOR HEALTH STATISTICS

DOROTHY P. RICE, Director

ROBERT A. ISRAEL, Deputy Director

JACOB J. FELDMAN, Ph.D., Associate Director for AnalysisGAIL F. FISHER, Ph.D., Associate Director for the Cooperative Health Statistics System

ROBERT A. ISRAEL, Acting Associate Director for Data SystemsROBERT M. THORNER, Sc.D., Acting Associate Director for International Statistics

ROBERT C. HUBER, Associate Director for ManagementMONROE G. SIRKEN, Ph.D., Associate Director for Mathematical Statistics

PETER L. HURLEY, Associate Director for OperationsJAMES M. ROBEY, Ph.D., Associate Director for Program Development

PAUL E. LEAVERTON, Ph.D., Associate Director for ResearchALICE HAYWOOD, Information Officer

DIVISION OF VITAL STATISTICS

JOHN E. PATTERSON, Director

ALICE M. HETZEL, Deputy Director

ROBERT L. HEUSER, M.A., Chief, Natality Statistics BranchJOSEPH D. FARRELL, Acting Chief, Programming BranchMABEL G. SMITH, Chief, Statistical Resources Branch

Vital and Health Statistics-Series 23-No. 3

DHEW Publication No. (PHS) 79-1979Library of Congress Catalog Card Number 79-11092

CONTENTS

Introduction

Summary of Principal Findings 2

Data and Methods 3

Trends in Breast Feeding Among Currently Married Women 4

Differentials in Breast Feeding............................................................................................................... 7Religion 9Race and Ethnicity 9Farm Background 9Geographic Region 9Education 9Occupation 10Poverty Level.................................................................................................................................... 10Employment 10Marital Status, Parity, and Mother’s Birth Cohort 10

Conclusions 12

References.............................................................................................................................................. 12

List of Detailed Tables 13

AppendixesL Technical Notes 27II. Definition of Terms in the National Survey of Family Growth.................................................. 36III. Items on NSFG Questionnaire Related to Breast Feeding 39

LIST OF FIGURES

1. Percent of ever-married women 15-44 years of age who breast fed their first child, by duration ofbreast feeding, year of first birth, and race: United States, 1973 2

2. Percent of ever-married women 15-44 years of age who breast fed their first or second child, byduration of breast feeding and order of birth: United States, 1973 3

3. Percent of ever-married women 15-44 years of age who breast fed their first child, by duration ofbreast feeding and employment between first and second births: United States, 1973 3

LIST OF TEXT TABLES

A. Percent of currently married women who breast fed their first or second child, by birth cohort ofmother and year of birth of child: United States, 1973 and 1965 4

B. Percent of ever-married women 15-44 years of age who breast fed their first or second child, byduration of breast feeding, birth cohort of mother, and year of birth of child: United States,1973 6

Percent of ever-married women 15-44 years of age with birth or more who breast fed their firstchild, by duration of breast feeding, year of first birth, and selected characteristics: Unitedstates, 1973

Percent of currently married women 1544 years of age with birth or more who breast fed theirfirst cndd, by duration of breast feeding, year of first birth, and selected characteristics: Unitedot3tS fio

SYMBOLS

Data not available----------__-

Category not applicable---------_

Quantity zero-----------_-.....

Quantity more than 0 but less than 0.05-- 0.0

Figure does not meet standards ofreliability or precision-

TRENDS IN BREAST FEEDINGAMONG AMERICAN MOTHERS

Charles Hirschman, Ph.D., Associate Professor, Department of Sociology, Duke University,and Gerry E. Hendershot, Ph.D., Division of Vital Statistics, National Center for Health Statistics

INTRODUCTION

In recent decades there was a substantial de-cline in the proportion of mothers who havebreast fed their children. The decline occurred inmost socioeconomic and cultural groups but was

greater among poor women, black women, andwomen with fewer years of education thanamong other women. In the early 1970’s thedecline may have leveled off or even reversed,especially in the West Region, although thatrecent change is not large enough to be statisti-cally significant.

The trends and differentials in breast feedingamong American women are analyzed in thisreport using information collected in the 1973National Survey of Family Growth and the 1965National Fertility Study. Trends in the percentof women who breast fed their first and secondbabies and the percent who breast fed them for3 months or more are examined in relation to

various socioeconomic and cultural character-istics of the mothers. Trends are examinedacross cohorts of women born in the same years,years of birth of their infants, and the surveyyears, 1965 and 1973.

Trends and differentials in breast feeding are

significant in several ways. Medical research hasfound that mothers’ milk may have health bene-fits for newborn infants, including short-termimmunity from some diseases and a nutritionalcomposition that reduces the incidence of bothmalnutrition and obesity in breast fed babies.1

Breast feeding also prolongs the period of postpartum amenorrhea, thus acting as a natural con-

traceptive, although for a short interval of un-certain length.24 Also, it has been suggestedthat women who breast feed have a lower riskof breast cancer mortality,5 but epidemiologicalstudies have not produced conclusive evidenceof that association.6’7 Various other physiologi-cal and psychological effects on mothers and in-fants which have been attributed to breast feed-ing are under scientific investigation.

In spite of the importance of breast feeding,little information is available about its preva-lence at the present time, trends in the recentpast, or differences among various groups. Theone systematic study with a nationally repre-sentative sample was published in 1974 byHirschman and Sweet8 with data from the1965 National Fertility Study. Hirschmanand Sweet documented a decline in breast feed-ing up to 1965 and examined its relationship toeducation, ethnicity, farm origin, and a numberof other socioeconomic and cultural character-istics of the mothers. Their analyses consideredbreast feeding of first-born infants for any dura-tion bv women who were married and livingwith their husbands at the time of interview. Inaddition to updating the Hirschman and Sweetanalyses to 1973, this report extends their find-ings by including second-bom infants and byconsidering variations in duration of breastfeeding. It also includes all women who havebeen married, whether or not they are currentlymarried.

SUMMARY OF PRINCIPALFINDINGS

Both the 1973 National Survey of FamilyGrowth (NSFG) and the 1965 National FertilityStudy (NFS) data show the dramatic decline in

the incidence of breast feeding in recent genera-tions of American women. Trends by birth co-

horts of women show that two-thirds of thewomen born in the second decade of this cen-

tury breast fed their first infant, but only one-

quarter of the women bom in the late 1940’sand early 1950’s did so. A similar trend is

observed in the period rates by year of birth ofinfant. Over 70 percent of first-born infants inthe 1930’s were breast fed, while less than 30percent in the late 1960’s and early 1970’swere breast fed (see figure 1). The declineleveled off in the early 1970’s, but it is too earlyto say if this foretells a rise in the rate, of breastfeeding.

The decline in long-term breast feeding-theproportion of mothers who breast fed their in-fants for 3 months or more-has been even more

precipitous. According to the 1973 NSFG, lessthan 10 percent of the mothers whose first child

byAW^l Breast fed duration

t^^^f Breast fed months

s iR g $ g3S

g a ^ -g S

YEAR OF FIRST BIRTH

Figure 1. Percent of ever-married women 15-44 years of age who breast fed their first child, by duration of breast feeding, year of first

birth, and race: United States, 1973

was born from 1966 to 1973 had breast fedtheir baby for 3 months or more, a substantialdecline from earlier periods. This indicates thatmore than two-thirds of the women who breastfed their infants in recent years had stopped bythe time the child was 3 months old, presumablyswitching to bottle feeding.

Second-born babies are considerably lesslikely than first-born babies to be breast fed.Among breast fed babies, however, second-bornbabies are more likely than first-bom babies tobe breast fed for longer durations (see figure 2).

The level and trend in breast feeding varieswidely across various socioeconomic and cul-tural categories. Several groups show a very highincidence of breast feeding, with about 50 per-cent breast feeding the first baby and withalmost no downward trend in the last 20 years.Included are women living in the West, womenwith 16 years or more of education, and womenwho have had jobs as professionals or managers.Among the groups that have experienced the

Not employedbetween first

and second births

Employedfirst and

second births

One birth only(no second birth)

Figure3. Percent of ever-married women 15-44 years of agewho breast fed their first child, by duration of breast feedingand employment between first and second births: UnitedStates, 1973

most precipitous declines in breast feeding levelsover the past two decades are black women,women with less than 12 years of education, andwomen who have never worked outside thehome. There are small differences in the levelsof breast feeding of first children among womenclassified by farm or nonfarm origins, current

poverty status, religion, and work experiencebetween the births of their first and secondchildren (see figure 3).

Figure 2. Percent of ever-married women 15-44 years of agewho breast fed their first or second child, by duration ofbreast feeding and order of birth: United States, 1973

DATA AND METHODS

The primary source of data for this study isCycle I of the National Survey of Family

Growth (NSFG-I). In NSFG-I, conducted in1973 by the National Center for Health Sta-tistics, interviews were held with a nationwide,area probability sample of 9,797 women aged15-44 years who had ever been married or whohad children of their own living in the house-hold. The focus of the survey was on fertility,family planning, and related maternal and childhealth topics. The NSFG extends the time seriesof earlier cross-sectional fertility surveys doneunder other auspices, particularly the 1965 and1970 National Fertility Studies. Interviewingwas done under contract with the NationalOpinion Research Center from July 1973 toFebruary 1974, centered on September 13,1973. Further details of the sampling design andother technical aspects of the NSFG are availa-ble in appendix I of this report.

In table A, data from the 1965 National Fer-tility Study are also used. The 1965 NFS wassimilar in content and structure to the 1973NSFG, but its sample consisted of 5,617 marriedwomen, with spouse present, born after July 1,1910. When comparisons are made between thetwo surveys, only those women who were cur-rently married in the 1973 NSFG are included.Further details on the sample and content ofthe 1965 NFS are available in a report by Ryderand Westoff.9

Throughout this study, the focus is ontrends and differentials in the proportion ofwomen who breast fed their babies, not the pro-portion of babies who were breast fed. Withthis focus, the findings presented in this reportshow the comparative frequency with whichmothers in different groups have breast fed theirinfants.

Because cumulative opportunity for breastfeeding increases with the number of births,which varies from woman to woman, it is im-portant to .compare the behavior of women withequal numbers of births. The approach in thisreport is to analyze breast feeding behaviorseparately for women when they had their firstchild and when they had their second. In each

/case the sample is defined accordingly; that is,the proportion who breast fed their first baby isbased on the population of women who had oneor more births, and the proportion who breastfed their second baby is based on the population

Table A. Percent of currently married women who breast fedtheir first or second child, by birth cohort of mother andyear of birth of child: United States, 1973 and 19651

Birth cohort of motherand year of birth of child

Birth cohort of mother

1951-59................................1946-50................................1941-45................................1936-40................................1931-35................................1926-304..............................1921-25................................1916-20................................1911-15................................

Year of birth of child

1971-73................................1966-70...............................,

1961-651956-60................................1951-55................................1946-505..............................1941-45................................1936-40................................1931-35................................

First child2

1973

38.4

25.625.737.743.249.647.1

29.128.338.043.148.858.9

1965

50.5

21.834.836.645.947.557.865.768.3

32.336.046.650.564.577.472.0

Second child3

1973

26.8

16.619.826.126.831.432.6

23.622.724.628.334.455.0

1965

37.6

13.824.125.930.833.440.049.462.8

22.727.533.241.353.765.975.7

’Data from the 196S National Fertility Study includewomen born after July 1, 1910; data from the 1973 NationalSurvey of Family Growth, Cycle I, include women 15-44 yearsof age.

Includes women with or more live births, except that in1973 babies who did not live with the mother for 2 months ormore are not included.

Includes women with 2 or more live births, except that in1973 babies who did not live with the mother for 2 months ormore are not included.

41929-30 for 1973 data.^lysO or before for 1973 data.

of women who had two or more births. In thisapproach, comparisons are made among womenwhose opportunity for breast feeding is similar,insofar as it is determined by number of chil-dren, so differences in breast feeding amongsocioeconomic groups can be attributed to fac-tors other than number of children.

TRENDS IN BREAST FEEDINGAMONG CURRENTLYMARRIED WOMEN

By linking the retrospective data from boththe 1965 NFS and the 1973 NSFG into one

time series, a much longer trend analysis is possi-ble than with either survey alone. In order to

compare the 1973 NSFG with the 1965 NFS,the NSFG sample has been restricted to cur-rently married women, the population repre-sented by the 1965 NFS. Table A shows the pro-portion of mothers who breast fed their first andsecond children, using both the 1965 NFS dataand 1973 NSFG data, according to birth cohortof women (i.e., the group of women bom in thesame period of years) and year of birth ofinfant.

In general, the discussion below directsattention to the larger differences and broaderpatterns which may be inferred with confidenceto reflect trends in behavior. There are severalreasons for exercising such caution in inferringtrends from these data. In particular, it shouldbe noted that the samples of the earliest cohortsin both surveys underrepresent women in thosecohorts who married relatively late in life. Thus,comparisons of those cohorts with later cohortsare affected by differences in average marriageage as well as differences in breast feeding be-havior. Also, comparisons of the 1965 NFS withthe 1973 NSFG may reflect not only differencesin breast feeding practice, but also some differ-ences in the surveys themselves-sampling vari-ability, minor differences in question wording,and changes in the population represented dueto aging or marrying. These aspects of the dataare discussed further in appendix I of thisreport.

Both the 1965 NFS a.nd\973 NSFG datashow a dramatic decline in the proportion ofwomen who have breast fed their children. Thedownward trend is evident for both first andsecond babies, for trends arranged by birth co-

horts of women, and by year of birth of infants.According to the 1965 NFS, 68.3 percent ofwomen in the 1911-15 birth cohort breast fedtheir first child, but this figure dropped to 34.8percent of women born in the early 1940’s. TheNSFG data show a continuing decline in breastfeeding among women bom in the late 1940’sand in the 1950’s. Only about one-quarter of thewomen in these recent cohorts had breast fedtheir first child.

The cohort trend in breast feeding’ of secondchildren is comparable to that of first births,

though at lower levels. For instance, almost two-thirds of the women in the 1916-20 birth cohortbreast fed their first child, but only one-halfbreast fed their second child. For the more

recent cohort of NFS women, born from 1941to 1945, over one-third breast fed their firstchild, but only one-quarter breast fed their sec-ond child. For the youngest cohorts in theNSFG sample, only about one-fifth of themothers breast fed their second child.

According to the 1965 NFS, the periodtrend, with data arranged by year of birth of theinfant, shows a similar decline in breast feeding,from over 70 percent of first-born babies in the1930’s to slightly above 30 percent in the early1960’s. The 1973 NSFG data show a similar de-cline, with a leveling off for the most recent

periods. In the early 1960’s, 38.0 percent offirst-born babies were breast fed; this dropped to

28.3 ’percent in the late 1960’s, but it roseslightly to 29.1 percent for the 1971-73 period.This rise is not statistically significant, but itmay suggest an approaching end to the long-term secular decline in the proportion ofmothers who breast feed their infants. (Seeappendix I for a discussion of standards of sta-tistical reliability used in this report.)

A comparable downward trend with a slightrecent reversal is also evident in the data for sec-ond births. The differential between first andsecond births noted in the cohort trends, withmothers being less likely to breast feed secondchildren, also is evident in the data arranged byperiods, or birth dates of infants. For the mostrecent periods, about 28-29 percent of mothersbreast fed their first child but only about 23-24percent breast fed their second child.

Table B shows the proportion of womenwho breast fed their infants, divided into twoduration categories-less than 3 months and 3months or more. Breast feeding of longer dura-tion is more likely to have the effects on motherand child previously mentioned-immunizationof the baby from some diseases, better nutritionof the baby, reduced risk of pregnancy for themother, and reduced risk of breast cancer forthe mother. Investigators do not know definitelythat breast feeding has all of these effects, andeven less is known about the amount of breastfeeding which may be necessary to produce an

Table B. Percent of ever-married women 15-44 years of age who breast fed their first or second child, by duration of breast feeding,birth cohort of mother, and year of birth of child: United States, 1973

Birth cohort of mother and yearof birth of child


1951-59.............................................................................1946-50.............................................................................1941-451936-40.............................................................................1931-35.............................................................................1929-30

Year of birth of child

1971-73.............................................................................1966-70.............................................................................1961-65.............................................................................1956-60.............................................................................1951-55.............................................................................1950 or before...................................................................

Duration of breast feeding

All Less thandurations j| 3 months

Perc

38.6

25.125.437.843.250.049.6

28.727.937.542.949.859.9

ent breast feefirst child1

25.6

21.118.725.427.631.329.4

21.619.725.229.531.727.9

3 monthsor more

eding

13.0

4.06.7

12.415.618.720.2

7.18.2

12.313.418.132.0

Duratic

Alldurations

Perce

27.0

17.819.426.226.731.934.4

23.522.124.729.134.256.7

on of breast feeding

Less than3 months

nt breast feeecond child2

17.0

13.714.217.116.119.818.8

16.415.215.418.519.727.1

3 monthsor more

ding

10.0

4.15.29.110.612.115.6

7.16.99.310.614.529.6

includes women with or more live births whose first baby lived with them for 2 months or more.Includes women with 2 or more live births whose second baby lived with them for 2 months more.

effect. It is estimated, however, that the contra-

ceptive effect of nursing is not likely to reducefertility unless breast feeding continues for at

least 3 months.4 For that reason, 3 months hasbeen used to divide long- from short-term breastfeeding in this study. Because long-term breastfeeding is more likely to affect mother andchild, trends in its incidence are especially sig-nificant.

Only the 1973 NSFG sample is used intable B. Unlike table A, in which previouslymarried women were excluded to make theNSFG sample comparable to the NFS sample,table B includes all women who had been mar-

ried, whether or not they were married at thetime of interview. Because of that difference insample coverage, the proportions who breast fedat all differ somewhat in tables A and B.

The figures in table B show a greater declinein the proportions reporting long-term breastfeeding than in the overall rate. Among firstbirths to the oldest cohorts of women in the1973 NSFG, those born in the late 1920’s andearly 1930’s, there was a nearly equal distribu-

tion of short-term and long-term breast feeding.For instance, about 50 percent of all women inthese cohorts breast fed their first child, andabout 2 out of 5 of them breast fed for 3months or more. Among the youngest cohortsof women, those born in the late 1940’s and1950’s, the overall rate of breast feeding hadbeen reduced by half to only 25 percent of allwomen, and only 1 out of 5 of them continuedbreast feeding for 3 months or more. For sec-ond children, the same downward trend was evi-dent, with a greater decrease in long-termbreast feeding than short-term breast feeding.

The pattern by year of birth of the childparallels that by birth cohort of the mother-asteep decline in the fraction of women whobreast fed their infants for 3 months or more.

Furthermore, the decline continues into the1970’s, with no sign of an increase in long-termbreast feeding for first children. (The small in-crease for second children is not statistically sig-nificant.) The trend is reaching such a low point,however, that further decreases must be verysmall; less than 10 percent of mothers in recent

years have breast fed their first or second in-fants for 3 months or more.

Two aspects of the trends in breast feedingdeserve emphasis-the large and rapid decline inbreast feeding during the last quarter centuryand the slowing and possible reversal of that de-cline in the most recent period. Full analysis ofthe causes and consequences of those trends isbeyond the scope of this report, but mentioncan be made of some factors which would beincluded in such an analysis.

The change from breast feeding to bottlefeeding was made possible by improvements inthe technology of bottle feeding, such as pre-mixed, water-soluble, powdered formula and thecollapsible, sanitary, disposable plastic bottle,which made bottle feeding convenient, safe, andinexpensive. Additional motivation of mothersto bottle feed was provided by their increasedopportunities in the workplace and elsewhereoutside the home. Responding to those oppor-tunities required more frequent separation ofmothers from their infants, which could be ac-

commodated better by bottle feeding thanbreast feeding. Also, bottle feeding was viewedas the "modern" feeding method, which encour-

aged its adoption.10As a consequence of the decline in breast

feeding, babies and mothers are now much more

dependent on bottle feeding technology, requir-ing that high standards of quality in productionand distribution be maintained. Also, to the ex-tent that breast feeding previously providedsome protection against unwanted and untimelypregnancy by its suppression of ovulation, thatprotection must now be foregone or provided bysome other contraceptive means. A further pos-sible consequence of the decline in breast feed-ing is an increased risk of breast cancer. Whilethe role of breast feeding in cancer etiology isnot now considered by leading researchers to bevery important,6’7 it is still under’s investiga-tion. l} So long as it is under investigation, someapprehension will remain about the effect ofthe decline in breast feeding on’cancer risks.Finally, the decline in breast feeding may haveconsequences for the emotional development ofchildren. In the opinion of some developmentalpsychologists, breast feeding encourages thekind of mother-child interaction which is neces-

sary for normal psychological development in

the infant. Of course, that interaction can andusually does occur where bottle feeding is prac-ticed. However, specialists are concerned thatthe caring behavior needed to substitute forbreast feeding may not be provided to all bottle-fed babies.! 2,! 3

The slowing and possible reversal of the de-cline in breast feeding, if confirmed by subse-quent studies, suggests that publicity given thebenefits of breast feeding in recent years, sup-ported by many medical specialists, may beaffecting the choices of mothers and their doc-tors. Also, the recent growth of organizations to

promote breast feeding in the United States maybe succeeding in redefining breast feeding. Nolonger is it denigrated as "old fashioned"; in-stead, it is acclaimed as "natural," making itspractice more socially rewarding.10

DIFFERENTIALS INBREAST FEEDING

As an aspect of child rearing, breast feedingmight be expected to vary among social and cul-tural groups that have different ideas about howmothers should care for their children, ideaswhich are taught to each new generation. Also,the amount of time and money available to vari-ous socioeconomic groups and the advice theyget from doctors, nurses, and other professionalsmight influence their breast feeding practices.In order to highlight some of these variationsamong groups, table C shows the proportions ofever-married mothers who breast fed their firstchild and the proportions who breast fed for 3months or more, according to the year of birthof the child and selected cultural, social, andeconomic characteristics.

The characteristics selected are religion, raceand ethnicity, farm background, geographicregion, education, occupation, poverty levelincome, and work experience. The first fourcharacteristics are established at birth or inchildhood for many Americans and indicatemembership in social groups which have somedistinctive customs. In comparing breast feedingpractices of women in these groups, there is an

implicit hypothesis that they are influenced bydistinctive breast feeding customs learned in thegroup.

Table C. Percent of ever-married 15-44 years of age with birth who breast fed their first child, by duration of breast feeding, year of first lurch, and selected charactemlics:United Slates. 1973

Selected characteristic

Religion

Race and ethnicity

Farm background

Geographic region

Education

Elementary school,8 years less..................

High school, 9-11 years.......High school, 12 yearsCollege, 13-15College, 16

Occupation

Professionals and managers..Sales and clerical workers....

Craftworkers, operatives,

and farmworkers.............

Poverty level

Below poverty income.........

200-299 percent..................

Employment between first

and second births

Employed between

Not employed between

Breast fed duration

All

38.6

34.940.325.244.4

38.935.743.1

41.537.6

32.237.037.649.1

48.134.7

33.346556.0

34.050.634.242.1

36.1

44.639.234.539.4

28.7

403

42.1

Year of first birtn

1971-73

28.7

23.529.420.748.2

30.211.4

19.3

24.329.8

23.126.521.1535

18.115.724.843.5S2.1

13.449.225.332.6

18.6

31.331.425.828.3

27.9

38.2

315

1966.70

27.9

27.328518.627.0

29.413.735.2

26.728.2

26.927.422.438.2

32.217.223.335.257.1

24.447.324.627.7

21.0

30.122.5

26.631.7

24.5

28.3

30.4

1961-66

37.5

38.736.040.948.4

38.923.839.1

32.939.1

30.637.435.149.5

40.128.732.250.469.2

33.848.835.139.5

33.2

34.537.335.240.0

28.4

40.8

37.4

1956<0

42.9

38.844.523.157.1

43.042.155.1

44.842.2

31.342.543.854.2

53.039.839347.650.2

47.451.639.244.5

40.9

53.346.538.041.3

38.7

43.9

43.0

1951-55

49.8

40354.717.442.8

48.859.1S7.7

44.447.6

41.543.357.454.7

62.349.544.757.0

45.8

43.654.543.356.2

53.5

63.154.145546.9

43.8

50.0

50.2

19SOorbefore

59.9

52.761.0

82.3

56.272.673.2

66354.5

63.159.463.950.5

63.160.454.669.755.6

62.161.459.166.7

53.4

67.466.955.754.6

67.8

54.7

62.4

Breast fed 3

All

13.0

10.813.810.617.2

12.516.720.1

19.010.9

10.110.015.416.1

25.0

13.38.614.421.6

16.419.29.114.3

14.2

21.113.510.312.3

7.8

123

16.0

Year of first birin

1971-73

7.1

537.00.015.6

7.61.53.7

6.313

6.96.34.7

13.6

2.15.5439.419.1

5.317.2

5.27.0

2.8

5.85.77.1

8.5

5.6

13.5

17.6

1966-70

8.2

7.68-2

12.69.7

8.53.911.2

8.38.1

11.05.06.1

12.0

12.2

2.75.79.825.4

8.920.25.86.3

53

10.93.65312.1

6.5

7.8

9.9

1961-65

12.3

10311.823.822.6

12.4

9.516.3

12.812.1

9.811.212.316.4

15.17.59.519326.4

17.417.610.3

12.2

13.714.410.312.0

103

9.0

14.9

1956-60

13.4

13.413.4"4.118.4

12.718.031.7

20.410.5

7.611.317.2

16.6

27.015.58.613.619.1

21.916.99.813.1

16.1

27.214.910.510.8

12.6

13.8

13.2

19S1-55

18.1

14.819.86.816.7

16.728.433.4

265

13.8

10.511.028-219.0

38.322.611.912.612.7

21.220311.024.1

22.6

34.5

23.015.713.2

17.8

15.5

19.2

1950before

32.0

16.335.2

*as

28.045.437.7

42.723.8

26.526.440.325.6

39.631.222.2S4.618.6

41.933.426.836.2

30.0

45.238.430.323.8

36.0

26.0

35.3

lThe Hispanic origin classification made independently of racial classification and includes of all racial groups.

The second set of characteristics-education,occupation, poverty level income, and workexperience-change during the lifetimes of mostAmericans and affect behavior not so muchthrough custom as through the opportunitiesthey open or close. The implicit hypothesis incomparisons involving these characteristics isthat breast feeding practice is a matter of choiceamong alternatives determined by each woman’scurrent socioeconomic situation.

The two types of characteristics and the hy-potheses on which they bear, which might becalled the "cultural" and "socioeconomic" hy-potheses, were identified by Hirschman andSweet8 in the work previously cited. They con-cluded from their analyses that neither hypoth-esis was sufficient in itself to explain observeddifferences in breast feeding, but both hypoth-eses and possibly others would be required fora full explanation. In this report, no attempt is

made to test these hypotheses formally, butthey have guided the selection of the variablesfor analysis and presentation.

Religion

,It will be noted in table C that among majorreligious groups, Protestants are somewhat morelikely to breast feed than Catholic women, and

Jewish women are least likely. For women inthe remaining category of religion "Other ornone," the proportions who breast fed wereabout the same as those for Protestants. Thedownward trend over the past decades is evidentfor both Protestants and Catholics, althoughProtestants show a small upturn (not statisticallysignificant) in the early 1970’s, while Catholicsdo not. The small samples of women in the Jew-ish and the "Other or none" categories in someyears preclude interpretation of their trends.

Race and Ethnicity

The overall downward trend in breast feed-ing is evident for all three racial and ethnic cate-

gories in table C-black, white, and Hispanic-but the decline for black women is greatest. Forthe earliest periods, the majority of blackwomen breast fed their first babies, and theywere considerably more likely than whitewomen to breast feed their babies for 3 monthsor more. But by the early 1970’s, only 11.4 per-cent of black women breast fed their first child,and only 1.5 percent breast fed for 3 months ormore, compared with 30.2 and 7.6 percent ofwhite women in the same categories. The declineof breast feeding among Hispanic women hasbeen steeper than that among white women butnot as precipitous as among black women. Amodest rise in breast feeding (not statisticallysignificant) in the 1970’s was present only forwhite women and was limited to those breastfeeding for less than 3 months.

Farm Background

The data in table C indicate that before the1950’s, women who had grown up on farmswere more likely than nonfarm women to breastfeed their children, especially for 3 months ormore. But from the late 1960’s, continuing intothe 1970’s, the differential has been reversed,

with women of nonfarm background having a

higher proportion breast feeding, although thedifferences are not statistically significant.Women with farm origins continued the down-ward trend into the 1970’s, while among womenwithout a farm background the proportion whobreast fed their first babies did not change sig-nificantly.

Geographic Region

There were only minor differentials by geo-graphic region for the earliest two periods in thetime series in table C. But as the secular declinein breast feeding began in the 1950’s and con-

tinued into the 1960’s, the West Region was

relatively unaffected. Over 50 percent of womenliving in the West Region breast fed their firstbaby in the early 1970’s, 13.6 percent for 3months or more-at least twice the proportionsin other regions of the country. From the late1960’s to the early 1970’s, an upward trend inbreast feeding was evident in the West Region,but not in any of the other regions, except for a

small, nonsignificant increase in the proportionof women breast feeding for 3 months or morein the North Central Region.

Education

For births occurring in 1950 or earlier, therewere small differences in breast feeding betweenthe educational attainment categories shown intable C, with a tendency for breast feeding to beless common among women who completedhigh school. In the late 1950’s, there emerged a

U-shaped relationship between breast feedingand education. Breast feeding was more likelyamong women with the least or most years ofeducation, while women of intermediate educa-tion were least likely to breast feed. The patterncontinued to change, and in recent years thetendency has been for breast feeding to be leastcommon among those with the least educationand most common among those with the most,reversing the pattern reported for babies bom 25years earlier. These changes may also be de-scribed in this way: In the past generation,breast feeding declined greatly among less edu-cated women, it also declined, but less dramati-cally, among moderately educated women; and

it changed hardly at all among college educatedwomen.

Occupation

Women are classified by occupation accord-ing to the last job they held. The few women inthe sample who had never worked outside thehome are classified as "never worked." Theother occupational groupings in table C are:(1) professionals and managers, (2) sales andclerical workers, (3) service workers, and (4)craft workers, operatives, and farmworkers. Itshould be noted that the occupation in the lastjob may not reflect exactly the economic activ-ity of the woman after the birth of her first or

second child, the periods for which her breastfeeding behavior is reported.

There were only small, nonsignificant differ-entials in breast feeding by occupation for theearlier periods. By the late 1950’s, a downwardtrend was evident for women in all occupationsexcept those who were "professionals andmanagers" and those who "never worked." Inthe "never worked" category, the decline beganlater, in the early 1960’s, and accelerated in thenext decade. Since the early 1950’s there is noevidence of a consistent downward trend in theproportion of "professionals and managers" whobreast fed their first babies, either at all or 3months or more.

Poverty Level

The index of poverty is based on the ade-quacy of annual family income relative to thesize and composition of the family. In additionto those below the poverty level, the index iscategorized in table C to show those whose in-come put them at 100 to 199 percent of thepoverty level, 200 to 299 percent, or 300 per-cent or more. Because the poverty index meas-ures economic status at the time of interview in1973 or 1974 and may not reflect accurately theeconomic circumstances at the time of first orsecond births, its use in analyzing the effects ofincome differences on breast feeding is limited.

Poor and near-poor women were more likelythan other women to breast feed babies born be-fore 1960, but breast feeding has declined morerapidly among poor and near-poor women, so

that in recent years there is less difference be-tween their practice and that of wealthierwomen. These trends parallel those observed foreducational groups, and probably have the samecauses, because educational attainment and in-come are closely related.

Employment

The last variable in table C contrasts womenwho worked between their first and secondbirths and women who did not. A residual cate-gory contains women who had not borne a sec-ond baby at the time of interview. For the mostrecent periods, this residual category containsmost of the respondents because too little timehad passed between the first birth and the inter-view for many second births to have occurred.

Since working soon after a birth would tendto reduce opportunities for breast feeding, itmight be expected that women who worked be-tween births would report less breast feeding.However, from the early 1950’s on there is noassociation between work and breast feeding ofthe first child. Women who did not work weremore likely than working women to breast feedduring the earliest period, but the reverse is true

for the most recent period, although the differ-ence is not significant. There are no significantdifferences in other periods. However, womenwho did not work were generally more likelythan working women to breast feed for morethan 3 months. It does not appear that the in-crease in labor force participation by mothers isthe sole cause of the downward trend in breastfeeding, because breast feeding has declinedamong both groups of women, those whoworked between births and those who did not.

Marital Status, Parity, and Mother’sBirth Cohort

This discussion of breast feeding differentialshas been limited to statistics on breast feeding offirst-born babies by ever-married women, ac-

cording to the year of the baby’s birth, as shownin table C. Table D shows statistics parallel tothose in table C, except that it includes onlywomen who were married at the time of the in-terview. There is little difference between com-

10

Table D. Percent of currently married 15-44 of age with birth who breast fed their first child, by duration of breast feeding, year of first birth, and selected character-istics: United States. 1973


Religion

Catholic...............................

Other

Race and ethnicity

Black...................................

Farm background

Geographic region

Education

Elementary school,

High school, 9-11 years.......High school, 12 years..........

College. 16

Occupation

Professionals and managers..Sales and clerical workers....

Craftworkers, operatives,

Poverty level

Below poverty income.........

Employment between firstand second births

Employed between

Not employed between

Breast fed any duration

All

38.4

34.340.226.542.9

38.832.541.9

40.737.6

31.537.236.749.5

45.634.533.346.556.3

35.351.1

33.841.7

35.5

47.239.334.439.0

28.9

40.3

41.8

Year of first birth

1971-73

29.1

22.830.620.745.5

30.78.817.7

23.830.5

25.327.321.4

51.9

16.118.524.841.752.9

14.548.824.934.2

19.5

37.631.226.228.6

28.0

39.2

35.6

1966.70

28.3

26.729.124.227.8

29.414.234.7

27.1

28.6

28.428.021.738.8

33.715.223.335.757.6

27.348.224.628.1

20.2

29.423.5

27.031.6

2G.O

28.2

30.0

1961-65

38.0

39.036.742.946.7

38.824.038.3

31.540.1

28.838.635.7

51.1

38.628.332.451.872.2

36.251.934.740.4

32.5

42.235.136.240.0

26.8

41.0

38.4

1956-60

43.1

40.044.332.358.9

43.339.453.3

44.942.4

31.1

42.443.854.9

49.739.740.248.752.9

47.153.439.143.9

41.1

53.549.137.242.1

39.3

44.5

42.8

1951.55

48.8

39.144.317.4

40.7

48.356.357.9

54.845.9

39.642.955.954.7

59.847.645.857.134.0

44.650.543.455.1

52.5

67.156.343.645.8

43.4

49.8

48.8

1950before

58.9

47.661.2

72.8

56.074.869.3

66.453.3

57.857.364.052.1

57.463.652.767.143.6

58.661.456.068.1

53.1

63.568.961.752.2

66.2

50.6

62.8

8reast red months

All

12.8

10.813.611.515.8

12.415.2

20.0

18.210.9

10.29.615.115.8

23.913.18.7

14.221.8

16.719.68.5

14.0

14.0

24.3

13.810.212.1

8.3

11.0

16.0

Year of first birth

1971.73

7.3

6.47.3

13.2

7.81.64.3

5.97.7

7.56.15.0

13.7

2.5

6.55,1

8.518.7

6.016.55.28.0

3.1

7.15.57.4

8.5

5.8

11.6

19.8

1966-70

8.7

8.48.6

16.59.6

8.9-4.2

12.5

8.88.7

12.35.56.4

12.2

14.62.55.9

10.425.0

10.221.25.7

7.2

6.4

17.13.56.212.1

7.4

7.9

10.3

1961-65

12.6

10.812.325.022.1

12.7

-7.1

15.8

12.412.6

8.711.813.017.1

15.47.79.8

20.027.3

18.918.510.211.6

12.3

19.514.610.511.8

12.3

7.5

15.9

1956-60

13.2

13.113.44.7

14.2

12.619.930.5

21.110.0

7.811.317.115.5

23.0

15.58.8

13.322.0

20.817.2

9.112.7

17.2

27.616.910.011.1

15.3

13.6

12.7

1951.55

16.8

14.218.36.814.9

15.9

26.133.5

24.812.9

11.28.0

27.417.6

36.421.1

12.010.95.8

23.519.29.1

22.9

22.6

37.524.314.912.1

14.9

12.7

18.6

1950before

31,8

13.V35.2

72.8

28.649.737.7

42.523.9

22.324.841.826.8

39.630.822.458.223.6

34.640.025.737.3

29.0

47.741.832.823.8

44.3

24.7

34.7

lTl1e Hispanic origin classification made independently of racial classincation and includes of racial

parable statistics in the two tables because mostever-married women were currently married, andbreast feeding experiences of currently marriedand previously married women are not verydifferent. The data on currently married womenare presented in table D so that readers maycompare them with the data previously pub-lished by Hirschman and Sweet,8 but becausethey differ so little from the data for ever-

married women, the detailed tables (tables 1-12)are presented for ever-married women only. Thedetailed tables (tables 5-8) also show differen-tials among the same socioeconomic and culturalgroups for second-born babies and for differentbirth cohorts of mothers. The differentials andtrends found in the detailed tables are similarto those included in the text but differ in somerespects.

n

CONCLUSIONS

Two general conclusions can be made fromthe findings in this study. First, even if breastfeeding has significant effects on individualmothers or children, its overall impact in con-

temporary American society is relatively minorbecause in recent years less than one-third ofmothers have breast fed their first child, and lessthan one-tenth have breast fed for 3 months or

more. The fractions are even lower for secondbirths. This means that the majority of Ameri-can infants are dependent upon bottle fed, com-

mercially prepared formulas or cow’s milk fortheir primary source of nutrition during theirearly months of life.

Second, the practice of breast feeding todayis most common among relatively advantagedwomen in society. Among women with collegedegrees and those who work in high-statuswhite-collar occupations, about one-half breastfed their first baby and almost one-fifth con-

tinued for 3 months or more. Whether thesehigher-than-average levels of breast feeding are

due to socioeconomic and cultural factors or

different medical advice cannot be ascertainedfrom these data.

REFERENCES

Ijelliffe, D. B., and Jelliffe, E. F. P.: Human milk,nutrition, and the world resource crisis, in Philip H.Adelson, ed.. Food: Politics, Economics, Nutrition andResearch. Washington, D.C. American Association forthe Advancement of Science, 1975, pp. 65-69.

âin, A. K., et al.: Demographic aspects of lacta-

tion and postpartum amenorrhea. Demography 7(2):255-271, May 1970.

âxena, P. C.: Breast-feeding: Its effects upon post-partum amenorrhea. Soc Biol. 24(1):45-51, Spring 1977.

^Rolland, R.: Bibliography (with review) on contra-

ceptive effects of breast-feeding. Bibliography Repro-duction 28(l):l-4,July 1976.

îtagawa, E. M., and Hauser, P. M.: DifferentialMortality in the United States: A Study in Socioeco-nomic Epidemiology. Cambridge, Mass. Harvard Univer-

sity Press, 1973.SCorrea, P.: The epidemiology of cancer of the

breast. Am Clin Pathol. 64(6):720-727, Dec. 1975.

âcMahon, B., and Brown, J.: Etiology of humanbreast cancer: A review. Journal of the American CancerInstitute 50(l):21-42,Jan. 1973.

îrschman, C., and Sweet, J. A.: Social backgroundand breast-feeding among American mothers. Soc Biol.

21(1):39-57, Spring 1974.SRyder, N., and Westoff, C.: Reproduction in the

United States. Princeton, New Jersey. Princeton Univer-

sity Press, 1971.^Jelliffe, D. B.: Community and sociopolitical con-

siderations of breast-feeding, in Breast-feeding and theMother, Ciba Foundation Symposium 45 (new series).New York. Elsevier/Excerpta Medica/North-Holland,1976. pp. 159-172.lllng,, R., and Ho,J. H. C.: Unilateral breast-feeding

and breast cancer. The Lancet 8029:124-127, July 16,1977.^Fraiberg, S.: Every Child’s Birthright: In Defense of

Mothering. New York. Basic Books, Inc., 1977.l^Rossi, A. S.: A biosocial perspective on parenting.

Daedalus 106:1-31, 1977.^National Center for Health Statistics: National Sur-

vey of Family Growth, Cycle I: Sample design, estima-tion procedures, and variance estimation, by D. K.French. Vital and Health Statistics. Series 2-No. 76.DHEW Pub No. (PHS) 78-1350. Public Health Service.

Washington. U.S.. Government Printing Office, Jan.1978. ->;^National Center for Health Statistics: Replication:

An approach to the analysis of data from complex sur-

veys, by P.J.McCarthy. Vital and Health Statistics. PHSPub No. 1000-Series 2-No. 14. Public Health Service.Washington. U.S. Government Printing Office, Apr.1966.

12

LIST OF DETAILED TABLES1. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child, by birth cohort of

mother and selected characteristics. United States, 1973 14

2. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child for 3 months ormore, by birth cohort of mother and selected characteristics: United States, 1973 15

3. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child, by year of first birth

and selected characteristics: United States, 1973 16

4. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child for 3 months ormore, by year of first birth and selected characteristics: United States, 1973 17

5. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child, by birth cohortof mother and selected characteristics: United States. 1973 18

6. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child for 3 months or

more, by birth cohort of mother and selected characteristics: United States, 1973 19

7. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child, by year of secondbirth and selected characteristics: United States, 1973 20

8. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child for 3 months ormore, by year of second birth and selected characteristics: United States, 1973 21

9. Number of ever-married women 15-44 years of age with birth or more, by birth cohort of mother and selected character-istics: United States, 1973 22

10. Number of ever-married women 15-44 years of age with birth or more, by year of first birth and selected characteristics:United States, 1973 23

11. Number of ever-married women 15-44 years of age with 2 births or more, by birth cohort of mother and selected character-istics: United States, 1973 24

12. Number of ever-married women 15-44 years of age with 2 births or more, by year of second birth and selected character-istics: United States, 1973 25

13

Table 1. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child, by birth cohort ofmother and selected characteristics: United States, 1973

Birth cohort of motherSelected characteristic

Total 1951-59 1946-50 1941-45 1936-40 1931-35 1929-30

All women...................................................... 38.6 25.1

Religion

Catholic.......................................................................... 34.9 23.7Protestant....................................................................... 40.3 25.5Jewish 25.2Other or none................................................................. 44.4 23.8

Race and-ethnicity

White.............................................................................. 38.9 27.1Black.............................................................................. 35.7 13.8Hispanic origin1 43.1 29.2

Farm background

Farm.............................................................................. 41.5 26.4Nonfarm......................................................................... 37.6 24.8

Geographic region

Northeast....................................................................... 32.2 19.6North Central................................................................. 37.0 24.6South.....;....................................................................... 37.6 18.2West 49.1 42.5

Education

Elementary school, 8 years or less.................................. 48.1 22.3High school, 9-11 years.................................................. 34.7 19.7.High school, 12 years..................................................... 33.3 26.7College, 13-15 years....................................................... 46.5 47.1College, 16 years or more............................................... 56.0 *25.0

Occupation

Never worked................................................................. 34.0 12.5Professionals and managers............................................. 50.6 38.5Sales and clerical workers............................................... 34.2 28.2Service workers.............................................................. 42.1 32.9Craftworkers, operatives, and farmworkers 36.1 15.7

Poverty level

Below poverty income.................................................... 44.6 32.8100-199 percent............................................................. 39.2 25.5200-299 percent............................................................. 34.5 21.8300 percent or more....................................................... 39.4 21.8

Employment between first and second birth

No next birth................................................................. 28.7 25.8Employed between births............................................... 40.9 22.2Not employed between births........................................ 42.1 24.3

25.4 37.8 43.2 50.0

24.225.5

34.3

26.613.133.0

24.325.7

19.323.322.538.8.

30.918.921.335.650.0

28.833.622.229.422.3

27.324.625.425.3

23.228.325.6

35.538.040.349.5

38.927.640.0

35.938.4

22.835.636.447.8

47.632.730.046.260.9

35.954.831.439.135.8

43.236.234.939.3

31.539.139.3

38.944.735.458.2

42.746.343.8

43.343.2

32.743.742.344.4

53.337.039.545.763.4

46.157.839.242.340.0

54.844.638.742.2

39.546.441.9

43.853.321.958.6

48.962.162.3

55.647.2

40.645.356.257.4

58.454.743.958.147.1

49.851.645.356.151.9

58.559.945.546.4

42.051.550.7

49.6

35.856.4

53.8

48.359.558.8

49.349.8

42.449.554.352.8

62.351.743.650.455.1

60.451.541.366.146.9

59.156.641.548.8

31.243.454.4

!The Hispanic origin classification made independently of racial classification and includes women from all racial groups.

14

Table 2. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child for 3 months or more,by birth cohort of mother and selected characteristics: United States, 1973


Total 1951-59 1946-50 1941-45 1936-40 1931-35 1929-30

All women...................................................... 13.0 4.0

Religion

Catholic.......................................................................... 10.8 *4.5Protestant....................................................................... 13.8 3.3Jewish 10.6Other or none................................................................. 17.2 *10.7

Race and ethnicity

White.............................................................................. 12.5 4.6Black.............................................................................. 16.7 *1.2Hispanic origin1 20.1 *4.4

Farm background

Farm.............................................................................. 19.0 *5.9Nonfarm......................................................................... 10.9 3.6

Geographic region

Northeast....................................................................... 10.1 *2.2North Central................................................................. 10.0 *2.2South............................................................................. 15.4 *2.3West............................................................................... 16.1 10.8

Education

Elementary school, 8 years or less.................................. 25.0 *3.8High school, 9-11 years.................................................. 13.3 3.6High school, 12 years..................................................... 8.6 4.0College, 13-15 years....................................................... 14.4 *6.6College, 16 years or more............................................... 21.6 *25.0

Occupation

Never worked................................................................. 16.4 *3.7Professionals and managers............................................. 19.2 *7.3Sates and clerical workers............................................... 9.1 4.6Service workers.............................................................. 14.3 4.9Craftworkers, operatives, and farmworkers.................... 14.2 *1.8

Poverty level

Below poverty income.................................................... 21.1 *4.8100-199 percent............................................................. 13.5 4.0200-299 percent............................................................. 10.3 *2.4300 percent or more....................................................... 12.3 5.7


No next birth................................................................. 7.8 2.8Employed between births............................................... 12.2 *1.7Not employed between births........................................ 16.0 10.5

6.7 12.4 15.6 18.7

6.56.8

8.6

6.75.4

15.7

7.16.6

6.14.66.1

11.2

13.65.24.49.8

15.9

11.010.34.87.07.4

10.85.85.47.2

4.97.78.3

10.212.6

*23.322.9

12.89.714.5

14.111.9

11.110.911.717.0

19.810.57.7

16.125.7

20.220.98.89.913.5

19.79.810.813.3

10.810.214.6

12.416.716.221.9

14.225.621.9

19.014.1

10.114.019.117.8

29.315.011.413.224.9

18.420.912.218.214.7

29.317.411.513.4

19.114.515.6

16.719.8*5.426.0

16.834.033.3

28.913.7

13.512.628.516.8

34.925.611.517.118.2

33.221.410.924.822.4

35.525.715.413.8

17.616.519.9

20.2

9.025.2

*19.4

19.425.044.2

32.115.2

12.613.029.826.8

36.222.614.421.518.8

*12.922.514.124.926.6

25.028.422.416.3

*8.518.122.5

iThe Hispanic origin classification was made independently of racial classification and includes women from all racial groups.

15

Table 3. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child, by year of first birthand selected characteristics: United States, 1973

Year of first birth

Selected characteristicTotal 1971-74 1966-70 1961-65 1956-60 1951-55

All women...................................................... 38.6

Religion

Catholic.......................................................................... 34.9Protestant....................................................................... 40.3Jewish............................................................................ 25.2Other or none................................................................. 44.4

Race and ethnicity

White.............................................................................. 38.9Black.............................................................................. 35.7Hispanic origin1 43.1

Farm background

Farm.............................................................................. 41.5Nonfarm......................................................................... 37.6

Geographic region

Northeast....................................................................... 32.2North Central................................................................. 37.0South 37.6West............................................................................... 49.1

Education

Elementary school, 8 years or less.................................. 48.1High school, 9-11 years 34.7High school, 12 years 33.3College, 13-15 years 46.5College, 16 years or more............................................... 56.0

Occupation

Never worked................................................................. 34.0Professionals and managers............................................. 50.6Sales and clerical workers............................................... 34.2Service workers.............................................................. 42.1Craftworkers, operatives, and farmworkers 36.1

Poverty level

Below poverty income.................................................... 44.6100-199 percent............................................................. 39.2200-299 percent............................................................. 34.5300 percent or more....................................................... 39.4


No next birth 28.7Employed between births............................................... 40.9Not employed between births 42.1

28.7 27,9 37.5 42.9

23.529.4

"20.748.2

30.211.419.3

24.329.8

23.126.521.153.9

18.115.724.843.552.1

13.449.225.332.618.6

31.331.425.828.3

27.938.231.9

27.328.5*18.627.0

29.413.735.2

26.728.2

26.927.422.438.2

32.217.223.335.257.1

24.447.324.627.721.0

30.122.526.631.7

24.528.330.4

38.736.040.948.4

38.923.839.1

32.939.1

30.637.435.149.5

40.128.732.250.469.2

33.848.835.139.533.2

34.537.335.240.0

28.440.837.4

38.844.523.157.1

43.042.155.1

44.842.2

31.342.543.854.2

53.039.839.947.650.2

47.451.639.244.540.9

53.346.538.041,3

38.743.943.0

49.8

40.954.7-17.442.8

48.859.157.7

44.447.6

41.543.357.454.7

62.349.544.757.045.8

43.654.543.356.253.5

63.154.145.546.9

43.850.050.2

The Hispanic origin classification was made independently of racial classification’and includes women from all racial groups.

16

Table 4. Percent of ever-married women 15-44 years of age with birth or more who breast fed their first child for 3 months or more,by year of first birth and selected characteristics: United States, 1973

Year of first birth

Selected characteristic1950 or

beforeTotal 1971-74 1966-70 1961-65 1956-60 1951-55

All women...................................................... 13.0 7.1

Religion

Catholic.......................................................................... 10.8 5.9Protestant....................................................................... 13.8 7.0Jewish............................................................................ 10.6Other or none................................................................. 17.2 15.6

Race and ethnicity

White.................................’............................................. 12.5 7.6Black.............................................................................. 16.7 *1.5Hispanic origini 20.1 *3.7

Farm background

Farm.............................................................................. 19.0 6.3Nonfarm......................................................................... 10.9 7,3

Geographic region

Northeast....................................................................... 10.1 6.9North Central................................................................. 10.0 6.3South............................................................................. 15.4 4.7West............................................................................... 16.1 13.6

Education

Elementary school, 8 years or less.................................. 25.0 *2.1High school, 9-11 years.................................................. 13.3 5.5High school, 12 years..................................................... 8.6 4.9College, 13-15 years....................................................... 14.4 9.4College, 16 years or more............................................... 21.6 19.1

Occupation

Never worked................................................................. 16.4 *5.3Professionals and managers............................................. 19.2 17.2Sales and clerical workers................................................ 9.1 5.2Service workers.............................................................. 14.3 7.0Craftworkers, operatives, and farmworkers.................... 14.2 *2.8

Poverty level

Below poverty income.................................................... 21.1 *5.8100-199 percent......................................................^...... 13.5 5.7200-299 percent............................................................. 10.3 7.1300 percent or more....................................................... 12.3 8.5


No next birth.................................................:............... 7.8 5.6Employed between births............................................... 12.2 *13.5Not employed between births........................................ 16.0 17.6

8.2 12.3 13.4

7.68.2

12.69.7

8.5*3.911.2

8.38.1

11.05.06.1

12.0

12.22.75.79.8

25.4

8.920.25.86.35.9

10.93.65.9

12.1

6.57,8

9.9

10.911.8

*23.822.6

12.49.5

16.3

12.812.1

9.811.212.316.4

15.17.59.5

19.926.4

17.417.610.311.112.2

13.714.410.312.0

10.99.0

14.9

13.413.4*4.118.4

12.718.031.7

20.410.5

7.611.317.216.6

27.015.58.6

13.619.1

21.916.99.8

13.116.1

27.214.910.510.8

12.613.813.2

18.1

14.819.8*6.8

*16.7

16.728.433.4

26.913.8

10.511.028.219.0

38.322.611.912.612.7

21.220.911.024.122.6

34.523.015.713.2

17.815.519.2

32.0

16.335.2

48.8

28.045.437.7

42.723.8

26.526.440.325.6

39.631.222.254.6*18.6

41.933.426.836.230.0

45.238.430.323.8

36.026.035.3

The Hispanic origin classification was made independently of racial classification and includes women from all racial groups.

17

Table 5. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child, by birth cohort of

mother and selected characteristics: United States, 1973


Total 1951-59 1946-50 1941-45 1936-40 1931-35

All women...................................................... 27.0 17.8

Religion

Catholic.......................................................................... 22.5 14.8Protestant....................................................................... 29.1 16.2Jewish 15.5Other or none................................................................. 32.6 *16.9

Race and ethnicity

White.............................................................................. 26.7 19.6Black.............................................................................. 28.9 -11.9Hispanic origin1 29.9 *6.5

Farm background

Farm.............................................................................. 31.9 18.2Nonfarm......................................................................... 25.2 17.7

Geographic region

Northeast....................................................................... 20.5 *12.4North Central................................................................. 24.6 14.8South............................................................................. 28.2 16.4West............................................................................... 35.2 17.1

Education

Elementary school, 8 years or less.................................. 32.0 *6.5High school, 9-11 years.................................................. 23.5 16.1High school, 12 years..................................................... 22.8 21.4College, 13-15 years....................................................... 34.5 61.1College, 16 years or more............................................... 44.2 *100.0

Occupation

Never worked................................................................. 24.8 *1.6Professionals and managers............................................. 37.1 43.5Sales and clerical workers............................................... 23.6 17.5Service workers.............................................................. 28.0 27.0Craftworkers, operatives, and farmworkers.................... 26.1 *11.4

Poverty level

Below poverty income.................................................... 31.6 20.8100-199 percent............................................................. 27.7 17.1200-299 percent............................................................. 22.6 10.5300 percent or more....................................................... 28.3 29.4

Employment between second and third birth

No next birth................................................................. 22.1 18.7Employed between births............................................... 30.4 *24.1Not employed between births 30.8 *8.1

19.4 26.2 26.7 31.9

16.021.00.0

21.1

20.213.222.3

18.919.5

16.215.616.630.6

18.711.218.533.947.7

*9.126.119.323.715.8

20.715.920.321.4

20.817.315.9

21.727.240.339.6

26.918.724.3

26.226.2

21.226.824.033.7

23.921.221.932.652.6

26.242.420.930.221.3

29.221.723.730.3

25.623.928.5

22.328.6

*15.640.1

25.626.430.5

29.425.7

16.525.528.925.4

30.222.823.827.547.4

29.937.624.026.324.0

33.532.319.526.3

21.628.629.5

27.624.9*4.428.2

30.443.541.5

40.827.5

22.426.738.239.7

43.334.124.442.131.9

36.637.136.830.036.9

41.138.127.029.5

20.841.633.6

l^he Hispanic origin classification was made independently of racial classification and includes women from all racial groups

18

Table 6. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child for 3 months ormore, by birth cohort of mother and selected characteristics: United States, 1973


Total 1951-59 1946-50 1941-45 1936-40 1931-35 1929-30

All women...................................................... 10.0 4.1

Religion

Catholic.......................................................................... 8.3 *2.6Protestant....................................................................... 10.8 *3.9Jewish 7.6Other or none................................................................. 10.3 *8.7

Race and ethnicity

White.............................................................................. 9.3 *3.3Black.............................................................................. 15.3 *7.5Hispanic origin1 15.3 0.0

Farm background

Farm.............................................................................. 13.5 *4.3Nonfarm......................................................................... 8.6 *4.0

Geographic region

Northeast....................................................................... 7.7 *4.9North Central................................................................. 9.7 *4.8South 11.4 *3.5West............................................................................... 10.5 *3.5

Education

Elementary school, 8 years or less.................................. 15.1 *1.0High school,9-11 years.................................................. 9.5 *1.8High school, 12 years..................................................... 7.2 *7.7College, 13-15 years....................................................... 11.1 *20.6College, 16 years or more............................................... 18.6 *100.0

Occupation

Never worked................................................................. 12.6 1.1Professionals and managers............................................. 14.8 0.0Sales and clerical workers............................................... 7.1 *4.9Service workers.............................................................. 9.8 *8.2Craftworkers, operatives, and farmworkers................... 11.6 *0.6

Poverty level

Below poverty income......................................;............. 14.3 *1.0100-199 percent............................................................. 10.7 *3.6200-299 percent............................................................. 7.0 *6.9300 percent or more....................................................... 10.2 *8.5


No next birth................................................................. 7.2 *3.9Employed between births............................................... 10.6Not employed between births........................................ 12.7 *7.4

5.2 9.1 10.6 12.1

3.76.3

*0.3

5.4*4.2*7.3

6.04.9

6.1*3.14.67.7

*5.2*2.74.48.1

30.4

*8.2*9.15.2

*2.75.7

6.44.24.86.2

5.3*4.25.4

7.29.0

*27.419.3

9.56.8

13.1

11.58.4

7.69.88.5

10.8

10.56.36.9

12.121.1

18.515.56.98.37.4

12.46.87.2

11.2

8.97.9

10.2

9.911.0*7.6

*10.1

9.321.217.4

11.110.3

5.311.614.19.3

16.110.97.6

10.219.8

*11.715.87.7

12.010.9

18.314.65.49.3

6.210.513.8

10.713.2

12.5

10.426.820.0

18.68.9

9.310.915.511.6

20.115.67.6

11.314.6

13.414.67.4

13.217.0

19.616.18.9

10.2

7.314.713.6

15.6

10.218.5

*2.8

15.022.049.2

20.513.5

12.513.518.618.5

25.420.112.013.713.5

29.017.09.3

15.626.8

30.320.712.713.3

12.915.417.3

l^he Hispanic origin classification was made independently of racial classification and includes women from all racial groups,

19

Table 7. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child, by year of secondbirth and selected characteristics: United States, 1973

Year of second birth

Selected characteristic1950 or

beforeTotal 1971-74 1966-70 1961-65 1956-60 1951-55

All women...................................................... 27.0

Religion

Catholic.......................................................................... 22.5Protestant....................................................................... 29.1Jewish 15.5Other or none................................................................. 32.6

Race and ethnicity

White.............................................................................. 26.7Black.............................................................................. 28.9Hispanic origin1 29.9

Farm background

Farm.............................................................................. 31.9Nonfarm......................................................................... 25.2

Geographic region

Northeast....................................................................... 20.5North Central................................................................. 24.6South 28.2West..............................................................................’. 35.2

Education

Elementary school, 8 years or less.................................. 32.0High school, 9-11 years 23.5High school, 12 years 22.8College, 13-15 years 34.5College, 16 years or more............................................... 44.2

Occupation


Poverty level




23.5 22.1 24.7 29.1

21.123.9

*24.731.5

24.912.321.8

24,0

23.3

23.125.117.530.9

14.59.8

20.338.849.8

*6.843.821.923.816.1

24.015.425.529.1

24.2

18.2

16.623.027.044,2

22.612.612.7

19.223.0

16.819.418.935.6

12.714.118.928.953.3

19.739.118.826.312.6

16.618.021.626.9

21.324.122.8

22.126.429.921.3

24.923.333.3

25.824.4

16.822.626.833.4

27.419.822.931.140.0

22.530.522.926.823.0

30.826.019.325.6

18.326.928.5

25.331.9

24.1

28.335.343.0

36.325.9

18.226.133.638.3

40.130.424.033.533.3

32.138.424.326.632.8

39.536.521.626.6

22.428,8

32.2

34.2

29.026.1

*14.141.4

32.349.138.4

42.630.0

33.229.938.334.4

36.832.628.451.645.3

31.237.230.834.637.8

46.540.624.732.4

23.640.434.4

56.7

42.159.8

*53.6

54.362.667.0

71.042.0

56.449.566.347.0

66.953.649.7*38.9

0.0

92.2*10.558.042.669.4

56.058.652.957.6

61.658.655.0

iThe Hispanic origin classification made independently of racial classification and includes women from all racial groups.

20

Table 8. Percent of ever-married women 15-44 years of age with 2 births or more who breast fed their second child for 3 months or

more, by year of second birth and selected characteristics: United States, 1973


Selected characteristic1950 orbefore

Total 1971-74 1966-70 1961-65 1956-60 1951-55

All women...................................................... 10.0

Religion

Catholic.......................................................................... 8.3Protestant....................................................................... 10.8Jewish............................................................................ 7.6Other or none................................................................. 10.3

Race and ethnicity

White.............................................................................. 9.3Black.............................................................................. 15.3Hispanic origin1.......................................................’...... 15.3

Farm background

Farm.............................................................................. 13.5Nonfarm......................................................................... 8.6

Geographic region

Northeast....................................................................... 7.7North Central................................................................. 9.7South 11.4West............................................................................... 10.5

Education

Elementary school, 8 years or less.................................. 15.1High school, 9-11 years 9.5High school, 12 years 7.2College, 13-15 years....................................................... 11.1College, 16 years or more............................................... 18.6

Occupation


Poverty level




7.1 6.9 9.3 10.6

6.57.7

*6.1

7.5"4.4*7.5

7.97.0

7.78.76.16.1

*2.9*0.17.39.5

19.4

*6.813.26.46.5

’4.9

*4.43.47.3

11.6

6.70.017.7

3.87.4

*19.616.1

7.0*4.8*3.3

7.66.6

6.26.05,5

10.7

*4.23.64.7

12.318.9

14.813.96.04.74.2

5.94.36.59.1

6.37.87.6

10.28.9

12.25.9

9.110.822.2

11.68.4

7.59.3

10.19.8

13.76.27.1

11.524.6

10.113.38.28.09.4

14.59.94.410.4

7.07.5

12.0

9.111.5

15.4

9.818.015.6

12.49.9

7.412.012.39.6

14.515.18.46.3

13.8

*5.417.45.0

12.715.2

15.216.56.78.5

7.211.311.9

14.5

11.516.2

-1.4

13.026.127.0

22.410.7

8.410.021.415.5

18.816.69.6

20.813.7

*15.615.711.914.318.0

25.420.69.811.4

12.214.215.3

29.6

24.230.8

23.3

21.549.451.1

36.322.8

34.229.037.612.8

46.424.414.728.9

0.0

70.76.5*9.825.942.8

48.920.229.526.0

45.028.028.4

^The Hispanic origin classification was made independently of racial classification and includes from all racial groups.

21

Table 9. Number of ever-married women 15-44 years of age with birth or more, by birth cohort of mother and selectedcharacteristics: United States, 1973


All women...................................................

Religion

Catholic.......................................................................Protestant....................................................................Jewish..........................................................................Other or none..............................................................

Race and ethnicity

White...........................................................................Black...........................................................................Hispanic origin1...........................................................

Farm background

Farm............................................................................Nonfarm......................................................................

Geographic region

Northeast.....................................................................North Central..............................................................South...........................................................................West.............................................................................

Education

Elementary school, 8 years or lessHigh school, 9-11 years................................................High school, 12 years...................................................College, 13-15 years.....................................................College, 16 years or more

Occupation

Never worked..............................................................Professionals and managers..........................................Sales and clerical workersService workers............................................................Craftworkers, operatives, and farmworkers..................

Poverty level

Below poverty income.................................................100-199 percent..........................................................200-299 percent..........................................................300 percent or more....................................................


No next birth...............................................................Employed between births............................................Not employed between births......................................


Total 1951-59

24,454

6,97816,084

398993

21,5252,6801,795

6,39618,058

5,0836,4617,9674,942

2,3555,243

11,6553,0992,101

1,1663,6709,8354,9654,816

2,6475,5786,3089,920

5,7107,00611,736

2,259

5281,592

5135

1.895350203

3911,868

304578906471

242772

1,103141

2

25873

820593515

388848586438

1,573288397

1946-50

Nurr

4,943

1,3823,211

55294

4,354546399

1,0313,912

9101,2671,784982

356972

2,598693324

234587

2,0781,050994

5481,1591,4711,765

1.9701,3651,609

1941-45

nber in thou

5,391

1,6573,428

73232

4,790554371

1,2784,113

1,2251,4351,5941,136

3711,0052,551847616

259910

2,245993985

5101,1591,5002,223

1,0001,7612,630

1936-40

sands

4,956

1,4443,245117150

4,398484408

1,4643,492

1,0521,3121,5371,055

5151,0662,339554482

174799

2,0491,046888

5431,1121,2692.032

4881,7052,763

1931-35

5,084

1,4743,346112152

4,465553?t8

1,6883,396

1,1311,3611,642949

6351,0672,253627501

202953

1,918922

1,088

543959

1.1612,420

5321,4123,140

1929-30

1,821

4941,261

3630

1,62319367

5441,277

460508503349

237360811236176

39349725361347

116343322

1,041

148476

1,197

The Hispanic origin classification made independently of racial classification and includes women from all racial groups.

22

Table 10. Number of ever-married women 15-44 years of age with birth or more, by year of first birth and selected characteristics:United States, 1973


All women...................................................

Religion

Catholic.......................................................................Protestant....................................................................Jewish..........................................................................Other or none..............................................................

Race and ethnicity

White...........................................................................Black...........................................................................Hispanic origin1...........................................................

Farm background

Farm............................................................................Nonfarm......................................................................

Geographic region

Northeast.....................................................................North Central..............................................................South...........................................................................West.............................................................................

Education

Elementary school, 8 years or less...............................High school, 9-11 years................................................High school, 12 years...................................................College, 13-15 years.....................................................College, 16 years or more

Occupation

Never worked..............................................................Professionals and managers..........................................Sales and clerical workersService workers............................................................Craftworkers, operatives, and farmworkers..................

Poverty level

Below poverty income.................................................100-199 percent..........................................................

200-299 percent..........................................................

300 percent or more....................................................


No next birth...............................................................Employed between births............................................Not employed between births......................................

Year of first birth

Total

24,454

6,97816,084

398993

21,5252,6801,795

6,39618,058

5,0836,4617,9674,942

2,3555,24311,6553,0992,101

1,1663,6709,8354,9654,816

2,6475,5786,3089,920

5,7117,00711,736

1971-74

2,973

8551,888

43187

2,668263235

5642,410

540824

1,084525

207482

1,517407361

181482

1,301539470

312749814

1,098

2,566125283

1966-70

Nu

5,875

1,7693,751

68287

5,184623467

1,1774,699

1,3071,4441,9011,224

4101,0732,959839595

291857

2,4341,2351,059

6021,3541,7262,193

1,8501,7622,263

1961-65

mber in tho

5,103

1,5753,178

98252

4,532513375

1,2643,839

1,1081,4011,608986

3541,1012,534682431

241777

2,153922

1,010

5341,1431,3502,075

6311,8572,614

1956-60

usands

4,988

1,3933,324100171

4,438499328

1,4443,544

1,0711,3921,4791,046

4711,0612,426573457

210809

1,995967

1,007

4991,2051,2612,023

3741,5753,039

1951-55

4,031

1,1172,760

8965

3,560443316

1,3002,731

8501,0501,279851

540933

1,815513229

164581

1,596877812

465755894

1,917

2081,1602,663

1950 orbefore

1,484

2691,184

30

1,14334175

648835

208349617310

3745934038628

79164357425458

235373262614

82528873

l^’he Hispanic origin classification was made independently of racial classification and includes women from all racial groups.

23

Table 11 Number of ever-married women 15-44 years of age with 2 births or more, by birth cohort of mother and selected

characteristics: United States, 1973


Selected characteristicTotal 1951-59 1946-50 1941-45 1936-40 1931-35 1929-30

Number in thousands

All women................................................... _18,891 || 696 3,033 4,409 4,471 4,583

Religion

5,583 155 930 1,348 1,319 1,360

P^nt"’"’’’’’’’’’’"’’’’’""’"’’"’"’"""’"""’’’:^ 12304 476 1,919 2,821 2,930 2,985Protestant..................................................

^ g3 101 107

Sh^none::::::::::::::::::;;:::::::::::::::::;::::;;:::::;::;:: 676 65 163 177 121 130

Race and ethnicity

16700 546 2,640 3,941 3,974 4,063White........................................................---

^ ggg 446 437 458

^n.ongini::::::::::::::::::::::::;:;::::::::::::;:::::;::^:.... 1.427 69 307 300 367 328

Farm background

5 162 134 654 1,037 1,321 1,500

Non?arm:::::::::::::::::::::;::::;::::::::::::l:::::::^^^^^:::^ 13:729 563 2,378 3,372 3,150 3,083

Geographic region

Northeast 4,029 104 568 975 942 1,021

No h ^ntrar""""""""""""""""""""""""".:: 5090 173 771 1,213 1,218 1,231

ôr h Central.......................................^ ^ ^ ^ ^ 369 ^S:::;:;::::;:::::::::::::;:::;;:::::::;::::............................. 3,927 156 673 925 941 895

Education

Elementary school, 8 years or less........;......,............... 1,991 99 271 332 473 606

^^srh^-l^.r---"--’----------"--:::::::;: ^ "6 JS 2,90044 2,^ 2,^Co e gears’::::::::::::::::::::::::::::.:.................. 2:350 21 369 667 49^ 571

College; 16 years or more............................................ 1,462 83 402 395 413

Occupation

901 95 195 234 155 187

Sâ^ndmanagers::::;:;:;;:::::::::::;::::::::::::::::::: 2,740 31 220 649 674 843

l^woTe"’0"’5 -::::::::: 3:923 Ô ^ ^ ^ ^Cra^woZrs operatives.’and’farm^^^^^^^^ 3,870 176 718 826 845 992

Poverty level

7 177 162 426 463 531 495

’OO^Snr"6""""’"""’""""’’""’"’’"’^ 4; 44 345 869 1,038 1,046

200"299pe cen """"""""""""""""""""............... 4964 128 1,012 1,287 1,186 1,062200-299 percent............................................. g 5300 percent or more. /,ûo


Nonextbirth............................................................... 7.936 580 2,014 2,124 1,429 1,264

Employed between births............................................ 3,621 40 388 852 ,050 1,004

Not employed between births...................................... 7,333 76 630 1,433 1.992 î5_________________________________1----- ------1------i------’------------

iThe Hispanic origin classification was made independently of racial classification and includes women from all racial groups.

24

1,699

4701,172

3620

1,53416557

5161,183

420483459336

210340748231169

35322695332315

100335289975

525286887

Table 12. Number of ever-married women 15-44 years of age with 2 births or more, by year of second birth and selectedcharacteristics: United States, 1973


Religion

Race and ethnicity

White...........................................................................Black

Farm background

Geographic region

North Central..............................................................South...........................................................................West.............................................................................

Education

College, 13-15 years.....................................................

Occupation

Poverty level

100-199 percent..........................................................

200-299 percent..........................................................



Total

18,891

5,58312,304

328676

16,7002,0141,427

5,16213,729

4,0295,0905,8453,927

1,9914,3478,7412,3501,462

9012,7407.4573,9233,870

2,1774,5444,9647,206

7.9363,6217,333

1971-74

2,573

8441,547

33149

2,316224230

4952,078

600646800527

166519

1,268347274

150371

1,120463469

231767819757

2,39843132

1966-70

Num

4,352

1,2922,800

63196

3,841449335

1,0513,301

9551,1621,322912

361873

2,090615413

209604

1,775933831

549942

1,3071,554

2,554593

1,204

1961-65

ber in thou

4,478

1,3622,838104174

3,970450352

1,1683,310

9781,2591,332910

4221,0342,186525311

219641

1,771873974

5171,1261,1231,712

1,4721,0911,915

1956-60

sands

4,549

1,3582,985

92113

4,068463286

1,3913,158

9891,2521,451856

4721,0072,109602359

179736

1.785943906

5251,0271,1481,849

1,0711,1172,360

1951-55

2,460

6431,741

3640

2,164290194

8161,644

445657744615

398737965259100

110369892543545

257535473

1,195

401623

1,436

1950 orbefore

479

84393

2

34013929

242237

63115195106

17217712235

3418

11417

146

9914795139

40153286

^The Hispanic origin classification was made independently of racial classification and includes women from all racial groups.

25

APPENDIXES

CONTENTS

I. Technical Notes.............................................................................................................................. 27Background 27Statistical Design 27Measurement Process 28Data Reduction 28Reliability of Estimates 28Nonsampling Error 29

The 1965 National Fertility Study 34Background 34Standard Errors 35

Differences Between NSFG and NFS Estimates 35II. Definition of Terms in the National Survey of Family Growth 36in. Items on NSFG Questionnaire Related to Breast Feeding 39

LIST OF APPENDIX FIGURES

I. Relative standard errors for aggregates of women, by race 30

II. Relative standard errors for percent of total and white women (base of percent shown in curve inthousands) 31

LIST OF APPENDIX TABLES

I. Approximate. standard errors for estimated percents expressed in percentage points for whiteand total women; 1973 National Survey of Family Growth 32

II. Approximate standard errors for estimated percents expressed in percentage points for blackwomen: 1973 National Survey of Family Growth 32

III. Standard errors for estimated percents for currently married white women and currently mar-ried women of all races based on data from the 1965 National Fertility Study 34

APPENDIX

TECHNICAL NOTES

Background

This report is one of a series of statistical re-

ports based on information collected from a

nationwide sample of women by the NationalSurvey of Family Growth conducted by the Na-tional Center for Health Statistics.

The National Survey of Family Growth(NSFG) utilizes a questionnaire to obtain demo-graphic information and information on fertil-ity, family planning, and health factors relatedto childbearing. As data relating to various sub-jects within these broad topics are tabulated andanalyzed, separate reports are issued. The pres-ent report is based on data collected in the firstcycle of the survey, which was centered on Sep-tember 1973.

The population covered by the sample forthe NSFG is women 15-44 years of age living inhouseholds in the conterminous United Statesat the time of interview who were ever marriedor had offspring living with them. The sampledid not include women living in institutions or

group quarters. Personal interviews were con-ducted by the staff of the National OpinionResearch Center (NORC), Chicago, beginning inJuly 1973 and ending in February 1974.

Statistical Design

The sampling plan for the survey was a mul-tistage probability design. Black households andhouseholds of all other races were selected atdifferent probabilities so that the sample was

composed of about 40 percent black women and60 percent women of all other races. The sam-

ple was designed so that tabulations could beprovided for each of the four geographic regionsof the United States.

The first stage of the sample design consistedof drawing a sample of primary sampling units(PSU’s). A PSU consisted of a county, a smallgroup of contiguous counties, or a standardmetropolitan statistical area as defined by theU.S. Bureau of the Census in March 1971. Thesecond and third stages of sampling were used toselect several segments (clusters of about 100dwelling units) within each PSU. A systematicsample of dwelling units was then selected fromeach segment. Each sample dwelling unit wasvisited by an interviewer who listed all house-hold members. If a woman 15-44 years of age,ever married or with offspring in the household,was listed as being in the household, an ex-tended interview was conducted. If more thanone woman in the household met the eligibilitycriteria, one of the women was randomlyselected for an extended interview.

Since the design of the NSFG was a complexmultistage probability sample, the derivation ofestimates involved three basic operations:

Inflation by the reciprocal of the probabilityof selection.-the probability of selection isthe product of the probabilities of selectionfrom each step of selection in the design(PSU, segment, listing unit, household, andsample persons within household).Nonresponse adjustment.-The estimateswere inflated by a multiplication of two fac-tors. The first has the number of samplehouseholds in a given PSU and stratum as itsnumerator and the number of householdsscreened in the PSU and stratum as its de-nominator. The second factor has as itsnumerator the number of’screened house-holds with an eligible woman of a specific

27

age and race class and PSU group, and as itsdenominator, the number of women actu-ally interviewed in the same age and raceclass and PSU. Screener response for thetotal survey was 89.8 percent and interviewresponse was 90.2 percent for the total sam-ple, yielding an overall response of approxi-mately 81.0 percent.

Poststratification by marital status-age-race.-I\ie estimates are ratio adjustedwithin each of 12 age-race cells to an inde-pendent estimate of the population for ever-married women. These independent esti-mates were derived from the U.S. Bureau ofthe Census Current Population Surveys of1971-73. The numbers of;single women withoffspring living with them were inflated bysteps 1 and 2.

All figures are individually rounded; aggre-gate figures are rounded to the nearest thousand.The sums of aggregates and percentages may notadd up to the total due to the rounding.

The effect of the ratio-estimating process isto make the sample more closely representativeof the population of women 15-44 years of age,living in households in the conterminous UnitedStates, and ever-married or with offspring livingwith them. The final postratification reducesthe sample variance of the estimates for moststatistics.

Descriptive material on the sampling designand estimation procedures may be found inanother report.14

Measurement Process

Field operations for the survey were con-ducted by NORC as agent for NCHS. Their re-

sponsibilities included pretesting the interviewschedule, selecting the sample, interviewing re-

spondents, and carrying out quality controlchecks. The questionnaire was pretested inNovember 1972, and subsequent smaller fieldtrials were held in March 1973. Interviewerswere trained for a week prior to fieldwork andhad their first few schedules reviewed thor-oughly. During the first part of the fieldwork,each interview schedule was reviewed for thecompleteness of certain key items and more in-

tensive review and followup were performed iferrors were discovered. Review and followupwere reduced to a sample of each interviewer’swork in the later part of the fieldwork. A 10-percent sample of all households with tele-phones was recontacted to verify the interviewand the accuracy of a few items. All of theseoperations were monitored by NCHS.

The parts of the interview schedule applica-ble to this report are reproduced in appendix III.The complete schedules are available uponrequest. Two different forms were used, one forinterviewing currently married women and theother for interviewing widowed, divorced,separated, or single women with their own chil-dren living with them. The two forms differmainly in wording when reference is made to thehusband; there are a few questions in eachschedule that do not appear in the other.

Data Reduction

Coding and keying were done by NORC andthe UJS. Bureau of the Census. Each coder’swork was systematically sampled for verifica-tion. Keying at the U.S. Bureau of the Censuswas performed on key-to-disk equipment pro-gramed to reject invalid entries. Each keyer’swork was systematically sampled for verifica-tion. The data were edited by the U.S. Bureau ofthe Census and NCHS to minimize internal in-consistencies. After editing, value entries were

imputed to cases with missing data on an item-by-item basis. No item with more than 15percent missing data was included in the imputa-tion. The imputed value entry for a case wasselected from a randomly chosen case withsimilar characteristics such as race, age, andmarital status, using a procedure known as "hotdeck" imputation.

Reliability of Estimates

Since the statistics presented in this reportare based on a sample, they may differ some-what from the figures that would have been ob-tained if a complete census had been taken usingthe same questionnaires, instructions, inter-viewing personnel, and field procedures. Thischance difference between sample results and a

complete count is referred to as sampling error

28

and is measured by a statistic called the standarderror of estimate. The relative standard error ofan estimate is obtained by dividing the standarderror of the estimate by the estimate itself andis expressed as a percentage of the estimate.Included in this appendix are charts and tablesfrom which the relative standard errors can bedetermined for estimates shown in this report.In order to derive relative errors which would beapplicable to a wide variety of health statisticsand which could be prepared at a moderate cost,a number of approximations were required. Asa result, the charts provide an estimate of theapproximate relative standard error rather thanthe precise error for any specific statistic. Thestandard errors were computed using a pro-cedure known as balanced half-sample repli-cation.15

The chances are about 68 out of 100 that anestimate from the sample would differ from a

complete census by less than the standard error.The chances are about 95 out of 100 that thedifference between the sample estimate and a

complete count would be less than twice thestandard error. In this report, numbers and per-centages which have a standard error that ismore than 25 percent of the estimate itself areconsidered "unreliable." They are marked withan asterisk to caution the user but may be com-bined to make other types of comparisons ofgreater precision.

In this report, sample statistics are comparedamong subgroups or across years, using thenormal deviate test at the 0.05 level of confi-dence. A statistically significant differenceamong comparable proportions of other statis-tics from two or more subgroups is one suffi-ciently greater than zero that a difference, ofthat size or larger would be expected in less than5 percent of repeated samples of the same sizeand type if there were no true difference in thepopulations sampled. If the observed differenceor a larger one could be expected in more than 5percent of repeated samples, one cannot be suf-ficiently confident to conclude that there is atrue difference in the populations. When an ob-served difference is sufficiently greater than zeroto be statistically significant, the true differencein the population is estimated to lie between theobserved difference plus or minus 2 standard

errors of that difference in 95 out of 100samples.

When two or more sample statistics are com-

pared and they have only small, statistically non-significant differences among them, they may bereferred to as the "same" or "similar." However,where a substantial difference observed is foundnot to be statistically significant, one should notconclude that no difference exists, but simplythat such a difference cannot be established with95-percent confidence from this sample. Ob-served differences that are described in terms

such as "greater," "less," "larger," "smaller,"etc., have been tested and found statisticallysignificant. Lack of comment in the text aboutany two statistics does not mean the differencewas tested and found not to be significant.

The standard error of a difference betweentwo comparative statistics, say the proportionwith characteristic M among black women com-

pared with white women, is approximately thesquare root of the sum of the squares of thestandard errors of the statistics consideredseparately, or calculated by the formula,

d=P, P,is

/p2y2V P

+ p2 ]/2/l ^"Pza.

where P^ is the proportion for one group, andPg the proportion for the comparative group,and Vp and Vp are the relative standard errors

of -Pj and Pg, respectively. This formula willrepresent the actual standard error quiteaccurately for the difference between separateand uncorrelated characteristics, although it isonly a rough approximation in most other cases.The relative standard error of various propor-tions can be estimated from figures I and IIand tables I and II for statistics based on theNational Survey of Family Growth.

Nonsampling Error

In addition to sampling error, the survey re-sults are subject to several sources of potentialnonsampling error, including interview nonre-

sponse, nonresponse to individual questionswithin the interview, inconsistency of responses

29

Figure 1. Relative standard errors for aggregates of women, by race

1009080706050

40

I- Ju

2UJ 20(-)OCLU

^ 10r 80 7OC 60: 5aj

QOC 3

<Qz. 2

<V)

^ -J-> .8

i; :.-J .5

^ .4OC

.3

.2

s^

^^^

*^

s^

s

>.t

-:,

’""s. "11^^

>/a

^’<y

1-0K.,

I.

^fe.::1-,;^ ^’,

ss^^,

^s.

^ ^-!l,-..

1^^s;

2 3 4 5 6 7 8 9A

10

5 6 7 8 91

100

2 3 4 5 6 7 8 9A 2 3 4 5 6 7 8 9A1,000

89^10,000

2 3 4 5 6 789

ESTIMATED NUMBER OF WOMEN (IN THOUSANDS)

Example of use of chart: An aggregate of 2 million women (on the scale at the bottom of the chart) of all races has a relativestandard error of 4.8 percent, or a standard error of 96,000 (4.8 percent of 2 million).

Figure II. Relative standard errors for percent of total and white women (base of percent shown in curve in thousands)

LUlU;oqLUj

occrUJ

Qcr<Qz<01

5i

908070

60

50

40

30

20

10987

6

5

4

3

2

.9

.8

.7

.6

.5

.4

.3

.2

^^^N^^

-^S.^^.^S

^

’^^

^~E

"S

*<

s

-^

^>

s,

^s

^

’S

<k

^

s

s

^^

S

s

s

s

^s

<

>

<

s

5;:::,-*!

>6

^ ’

,-;

’i,

: -100

!,;’S200, ^300

^500’.,,

1,000

’’.^ 2,000

i, 3,000’

:: 5.000

.::: 10,000’’’ij,,. 20,000

’. ’<30,000

perce

S

s

s.

"s.-<-’s

s^\

^

nt

S

SsS

-^1

s--%

ss

tho

s

s

^s-

s\

s

ss

usa

S,

Ss

S;

ss

<

s

nc

s

s

s

*s

\

s

S\

Is) : ::

^-

\ ^\\^, ^^=^=?N^-J,

^-^-S--

^^\’>S>.--S-\ ^S

^s^S\^--^

s

^’IV -^

2 3 4 5 6 7 8 9A 2 3 4 5 6 7 8 91

10 100

ESTIMATED PERCENT

3 4 5 6 7 8 91

1,000

Example of use of chart. An estimate of 10 percent (from the scale at the bottom of the chart) ofa population subgroup of million women (fifth curve from the top) has a relative standard error of20.0 percent, or a standard error of 2.0 percent (20.0 percent of 10 percent).

to individual questions, respondent error or mis-reporting, and errors of recording, coding, or

keying by survey personnel. It is impossible tomeasure the extent of nonsampling errors accu-rately. Although some useful approximate meas-

ures can be made of some types of nonsamplingerror, the survey must rely upon several qualitycontrol procedures and other methods incorpo-rated into the survey design to minimize non-sampling error.

31

Table I. Approximate standard errors for estimated percents expressed in percentage points for white and total women: 1973 NationalSurvey of Family Growth

Base of percentage

100,000..............................................................................................................500,000..............................................................................................................1,000,000...........................................................................................................3,000,000...........................................................................................................5,000,000...........................................................................................................7,000,000...........................................................................................................10,000,000.........................................................................................................

Estimated percent

2 or98

3.01.30.90.50.40.30.3

5 or95

4.62.11,5

0.80.60.50.5

10 or

90

6.42.82.01.20.90.80.6

20 or80

8.53.82.71.51.21.00.8

30 or70

9.74.33.11.81.41.21.0

40 or60

10.44.63.31.91.51.21.0

50

10.64.73.31.91.51.31.1

Table II. Approximate standard errors for estimated percents expressed in percentage points for black women: 1973 National Surveyof Family Growth

Base of percentage

5,000..................................................................................................................’10,000................................................................................................................50,000................................................................................................................100,000..............................................................................................................300,000.............................................................................................................;500,000..............................................................................................................700,000..............................................................................................................1,000,000...........................................................................................................

Estimated percent

2 or

98

7.95.62.51.81.00.80.70.6

5 or95

12.38.73.92.71.61.21.00.9

10 or90

17.012.05.43.82.21.71.41.2

20 or

80

22.616.07.15.12.92.31.91.6

30 or70

25.918.38,2

5.83.32.62.21.8

40 or60

27.719.68.86.23.62.82.32.0

50

28.320.08.96.33.62.82.42.0

Interview Nonresponse.-Interview nonre-

sponse, or the failure to obtain whole interviews,arises from several sources-incomplete listing ofhouseholds for the sampling frame, inability to

screen all sample households for eligible re-

spondents, and inability to complete a full inter-view. Completeness of listing cannot be testeddirectly as it requires an independent, accurate

accounting of the households that should havebeen listed. In the NSFG, listing accuracy wastested at the time of screening by use of the"half open interval" check for missed house-holds; i.e., at designated sample households, theinterviewer was required to check for dwellingunits between the sample household justscreened and the next listed dwelling unit. Thisprocedure resulted in the addition of 781 missedunits or an additional 2.4 percent to the originalsample of dwelling units to be screened.

Of the original sample of 32,818 dwellingunits to be screened, 3,820 were found to bevacant, not dwelling units, or group quarters. Ofthe remaining dwelling units, 9.7 percent were

not successfully screened. This included 2.3 per-cent refusals to have the household memberslisted; 1.6 percent with language problems, ill-ness, or otherwise unavailable in the field period;4.6 percent where no one could be found athome; and 1.1 percent for other reasons such as

refused access to the unit.Of the 26,177 households for which screen-

ing was completed, 10,879 were found to con-

tain an eligible respondent. However, interviewswere not completed in 9.8 percent of these casesbecause of refusals by the eligible respondents(5.0 percent); language, illness, and relatedproblems (2.0 percent); and no contact afterrepeated calls (2.7 percent).

The nonresponse adjustment for interviewnonresponse described above imputes to non-

responding dwelling units and women the char-acteristics of similar respondent dwelling unitsand women.

Item nonresponse.-Nonresponse to individ-ual questions (item nonresponse) was less than2 percent for about half (51 percent) of the

32

items. Item nonresponse occurred when the per-son refused to answer the question, when theperson did not know the answer to the question,when the question was erroneously not asked orthe answer not recorded by the interviewer, andwhere the answer was uncodeable. For 37 per-cent of the items, nonresponse was between 2.0and 10.0 percent. For the remaining 12 percentof the items, nonresponse was greater than 10percent of persons eligible to answer the items.Half of these high nonresponse items were con-centrated in two areas-detailed income ques-tions and questions about the reasons for switch-ing from one contraceptive method to another.The remaining high nonresponse items were gen-erally those asked of small numbers of persons.

The amount of missing data or imputedvalues for various items will usually be shownwith the definition in appendix II, especiallywhere it is substantial. Some illustrative itemswith their associated nonresponse rates are: thenumber of children ever bom (parity) (nomissing data), intentions about having anotherchild (0.7 percent), whether contraception wasstopped in order to become pregnant (1.9 per-cent), highest grade of school attended (0.1 per-cent), and total family income (6.8 percent).

For most items an adjustment for missingdata values was made by one of four imputationprocedures. In order of frequency employedthey were: (1) "hot deck", imputation, (2) im-putation from a sorted file, (3) editing fromother data within the same case, and (4) alloca-tion based on technical judgments.

"Hot deck" imputation refers to a procedurein which the file is first randomized. Next amatrix is created for values of items (e.g., race,age, and marital status) judged to be correlatedwith the item to be imputed (e.g., number oftimes married). A reasonable "cold deck" value(e.g., 2 married twice) is assigned to each cellof the matrix in case the first file record withthe given characteristics has missing data. Therandomized file is processed and each record isidentified as belonging to one cell of the matrix(e.g., white, age 25-29, currently married). Theitem to be imputed is checked: if it is blank-not applicable (e.g., not married before), it isignored; if it has a missing data code, the codein the matrix is placed in the record. If it has anacceptable code, that code replaces the code

already in the matrix, and it remains in thematrix until another record with the same char-acteristics and a known code is encountered.This insures that the probability of a code beingassigned to a record with missing data is thesame as the probability of that code occurringamong records with the same characteristicsbut with known data.

For imputation from a sorted file, the rec-ords are first sorted by selected characteristics(e.g., marital status, race, and age) so that thefirst group of records would be currently mar-ried black women aged 15-19, the second groupwould be currently married black women aged20-24, etc. An initial value is assigned for theitem to be imputed-(e.g., 4 [tubal ligation] fortype of sterility)-and for any item dependentupon the item to be imputed-(e.g., 9 [notascertained] as to whether the operation was forcontraceptive reasons). The ordered file is proc-essed and each record is checked. If the item tobe imputed is blank-not applicable, it is ig-nored; if it has a known code, it and its depend-ent items would replace the existing set ofvalues; if it has a missing data code, it and itsdependent items would be changed to the presetvalues above. This procedure insures that the im-puted code is reasonable for the ordering charac-teristics and that the probability of assignmentis the same as in the population in general. Therewill be some bias, however, as the boundariesbetween groups are crossed.

Where sampling error affects the precisionof survey estimates, nonsampling error intro-duces bias. Imputation procedures reduce thisbias to the extent that the assumptions aboutthe relations between respondent and nonre-spondent characteristics are true. But theamount of remaining bias, if any, cannot bemeasured. Therefore, stringent quality controlprocedures were introduced at every stage ofthe survey, including the check on completenessof the household listing mentioned earlier, theextensive training and practice of interviewers,field observation of interviewers, field editing ofquestionnaires, short verification interviews witha subsample of respondents and missed house-holds, verification of coding and editing, anindependent recede of a sample of question-naires by NCHS, keypunch verification, and anextensive computer "cleaning" to check for

33

impermissible codes, missing data, and responseinconsistencies. One source of bias that can beevaluated through special studies but cannot becontrolled is respondent error, whether delib-erate or’unwitting. In this as in other surveys,the data are subject to problems of accuraterecall and of the stability of respondents’ viewsfrom one time to the next.

Age-at-marriage bias.-i.t should be notedthat among birth cohorts only those memberswho have been married or have children livingwith them are included in these samples. Sincemost American women are married before age25, the women interviewed at ages over 25 are

quite representative samples of all women intheir cohorts. But women interviewed under age25, and especially under age 20, are not repre-sentative of all women in their cohorts becausethe survey does not include the numerouswomen in their cohorts who had not yet marriedat the time of interview. In other words, thesamples of women in recent cohorts, the youngwomen, include only early-marrying women,while the samples of women in earlier cohorts,the older women, include both early- and late-marrying women. In comparing recent andearlier cohorts, therefore, groups are being com-

pared which have different age-at-marriagecompositions. To the extent that age at marriageis associated with breast feeding practice, com-parisons between cohorts are affected by differ-ences in age-at-marriage composition in additionto any effects of historical trends in breast feed-ing practice. This kind of problem is inherentin a survey such as the National Survey ofFamily Growth, which reconstructs past events

from reports of a sample of women interviewedat one point in time. It cannot be solved com-pletely and should be kept in mind by the readeras a caution to the interpretation of comparisonsbetween cohorts.

THE 1965 NATIONAL FERTILITYSTUDY

In this report, NSFG data are sometimescompared to data from the 1965 National Fer-tility Study (NFS).

Background

Like the 1973 NSFG, the 1965 NFS col-lected information on fertility and family plan-ning. The target population consisted of cur-rently married women born since July 1, 1910,who were living with their husbands and re-siding in the conterminous United States. Thesurvey was conducted by the Office of Popula-tion Research at Princeton University and spon-sored by the National Institute of Child Healthand Human Development. National Analysts,Inc., of Philadelphia drew the area probabilitysample, conducted the interviews, edited andcoded the questionnaires, and prepared the basicdata file.

The interview completion rate in the study,defined as the number of successfully completedinterviews divided by the estimated number ofwomen eligible to be interviewed, was 88 per-cent. Of the 12 percent not interviewed, sometwo-thirds, or 8 percent, were classified as"refusals;" the remainder were cases of the

Table Standard errors for estimated percents for currently married white women and currently married women of all races based ondata from the 1965 National Fertility Study

Number of women in thousands

500.........................................................................1,000......................................................................................2,000..........................................................5,000.....................................................................10,000.........................................................................20,000..........................................

Estimated percent

5 or95

2.41.71.20.90.70.6

10 or90

3.42.81.71.20.90.8

20 or80

5.03.22.41.61.31.1

30 or70

5.13.72.71.81.51.2

40 or60

5.44.02.82.01.51.3

50

5.54.02.92.01.61.3

34

respondent not at home despite calls and othermiscellaneous reasons. Interviews were com-pleted with 5,617 women. Further discussion ofthe design and conduct of the survey is found inthe study report by Ryder and Westoff.15

Standard Errors

Standard errors for the 1965 National Fer-tility Study are measures of sampling variabil-ity, the variation that occurs by chance becausea sample of units is surveyed rather than the en-tire population. The chances are about 68 out of100 that an estimate from the sample woulddiffer from the population value by less than 1standard error, about 95 out of 100 that the dif-ference would be less than twice the standarderror, and about 99 out of 100 that the differ-ence would be less than Wz times the standarderror.

The contractor for the 1965 survey pro-duced tables of approximate standard errors forpercentage estimates from which table III in thisappendix was derived.

DIFFERENCES BETWEEN NSFGAND NFS ESTIMATES

It will be noted that the NFS and NSFGdata for the same birth cohorts and years ofbirth yield somewhat different estimates of theproportions breast feeding. Several factors maycontribute to these differences. First, the dataare from probability samples, and two probabil-ity samples drawn independently from the samepopulation will produce different estimates ofthe same statistics, within theoretically knownlimits. Most of the differences between estimates

of the same statistics from these two surveys areso small that it is likely that they resulted fromchance sampling factors.

Second, there were small differences in thequestions on breast feeding asked in the twosurveys. The 1965 NFS simply asked women ifthey had breast fed the baby, while the 1973NSFG asked if they had breast fed the baby "atall." It is possible that women with identicalbreast feeding experience would answer thosetwo questions differently. Also, in 1965 thequestion about breast feeding was asked aboutall babies, but in 1973 it was asked only aboutbabies who lived with the mother for 2 months;thus, breast feeding of babies who died or wereseparated from their mothers within 2 monthswas included in the 1965 survey but not in the1973 survey.

Finally, the 8-year interval between the sur-veys affects estimates for the same birth cohortsand years of birth. The 1965 sample of a birthcohort represents women who were mothersat that time. By 1973, more women from thesame cohort had become mothers, making themeligible for the sample. Thus, the 1973 samplefrom a birth cohort represents more mothersthan the 1965 sample from the same cohort; thelarger 1973 sample might also be somewhat dif-ferent, on the average, from the 1965 sample.Comparisons between the two surveys for thesame years of babies’ births are affected in a re-lated way. Some of the women who reported in1965 on breast feeding of babies bom in a par-ticular period would have become too old (45years of age) to be included in the 1973 survey.Thus, the 1973 sample of women reporting on

babies born in that period would be somewhatsmaller and possibly somewhat different, on theaverage, than the 1965 sample reporting on thesame birth period.

0 00

35

APPENDIX II

DEFINITION OF TERMS IN THE NATIONAL SURVEY OF FAMILY GROWTH

Breast feeding.-for each of her childrenwho lived with her for at least 2 months afterbirth, mothers were asked, "When (CHILD) wasan infant, did you breast feed him/her at all?"If so, women were then asked, "How manyweeks old was he/she when you quit breast feed-ing him/her altogether?" Women who answeredthe first question affirmatively were consideredto have breast fed the child regardless of thelength of time reported in the following ques-tion. Reporting on breast feeding was very com-

plete; for instance, breast feeding status of thefirst child was ascertained for more than 99 per-cent of respondents.

Birth cohort.-A birth cohort is a group ofwomen born in a specified calendar period. Be-cause the information for classifying women inbirth cohorts comes from a sample of womenwho were married, previously married, or singlewith children of their own living in the house-hold, it may be a biased sample of all womenborn in a specific period. Women who were bornin a specific period but who were never marriedand without children at the time of the inter-view would not be represented in the birth co-horts constructed from these data. This bias isnegligible for early birth cohorts (older women),because very few of them had not married bythe time of interview; but the bias is significantfor recent birth cohorts, because many of themhave not yet married or borne children and arenot represented in the sample. The effect oftheir omission is a relative overrepresentation ofearly-marrying women in the recent birthcohorts constructed from the sample data.

Marital status.-Marital status was a criterionof sample selection. The NSFG sampled womenwho were currently married at the time of inter-view, had ever been married, or had never been

married but had offspring (i.e., children born tothem) in the household. Current marital statuswas recorded in seven categories in response tothe question, "Is (PERSON) now married,widowed, divorced or annulled, separated, or hashe/she never been married?" The seven cate-

gories in which answers were recorded were mar-ried, informal union, widowed, divorced or

annulled, separated, single with own children,and never married. Women in the last categorywere not eligible for the survey.

Married women include those who are

legally or formally married whose husbands are

living in the household or are temporarily absenton business, illness, vacation, etc., and thosewho are informally married or "living together"with a male partner whose usual residence is thesame household. Women currently in informalunions were reported separately but are too fewto be separately classified for analytical pur-poses. Information on informal unions was ob-tained only if volunteered by the respondent inthe course of listing household members andtheir relationship to the head of the household.

Divorced women include those legally sepa-rated from their former spouses by a legal decreeof divorce or annullment and free to remarry.While those legally separated but without free-dom to remarry belong in the later category ofseparated, there was no direct question in theinterview to establish the issue of freedom to

remarry with certainty. The term divorce is pre-sumed to refer most generally to "absolute"decrees.

Widowed women are those previously mar-ried women whose most recent spouses aredeceased.

Separated women are those legally or in-formally separated from their former spouses.

36

Included here would be cases of desertion, in-formal separation by mutual agreement, andlegal separations in which the partners are notfree to remarry.

Single with own children is a category ofwomen who have begun their childbearing andhave never been married. However, some ofthese women were probably missed in the surveybecause this category was not identified by adirect question. These are single women whohave one or more children born to them andliving with them in the household. Single womenwho gave a child up for adoption or who boardthe child elsewhere and those who have not hada live birth are not included in the survey.

Religion.-Women were classified by religionin response to the question, "Are you Protes-tant, Roman Catholic, Jewish, or somethingelse?" In addition to the three major religiousgroupings, two other categories-other andnone-were used. Since the category of Protes-tant includes numerous individual denomina-tions, these respondents were further asked to

identify the denomination to which theybelonged. Those who answered "other" to theoriginal question and then named a Protestantdenomination were then included with theirown groups. Although specific denominationalnames were obtained and recorded, the numbersof cases for most denominations were too few toproduce reliable estimates, so they have beencombined in larger categories. Data on religiousdenominations were reported for all but 26respondent cases, more than 99 percent, andthese few cases were imputed.

Race and ethnicity.-Women were classifiedas white, black, or "other races" according tothe interviewer’s observations at the time ofinterview. Agreement between this classificationand the respondent’s own reports of ethnicorigin, also obtained in the interview, was veryhigh; for instance, of those classified as "black"by interviewer observation, 100 percent re-

ported their ethnic origin as at least partly"black, African, or Negro"; and of those whoreported their ethnic origin as "black, African,or Negro," 96 percent were classified as "black"by interviewer observation. Race was imputedfor 10 cases.

Information about ethnic origins was ob-tained for the woman by asking, "What is your

origin or descent?" Persons were classified asbeing of Hispanic origin if any of the followingresponses were given: Mexicano, Chicano, Mexi-can American, Puerto Rican, Cuban, Hispano, orany other Spanish origin or descent. Personswho did not give any of those responses wereclassified as being not of Hispanic origin. Personsmay have more than one origin or descent, andmultiple responses to the questions were re-

corded. However, any of the responses listedabove resulted in classification of the person asbeing of Hispanic origin regardless of any otherresponses which may have accompanied it. Itshould be noted that in this report the classifi-cations of race and ethnicity are independent;each ethnic category may include persons of allraces, and each racial category may include per-sons of all ethnic groups. Ethnicity was reportedfor more than 99 percent of respondents.

Farm background.-Women were asked,"When you were growing up, that is, betweenthe ages of 6 and 16, did you live on a farmmost of the time (half of the time or more)?"Women who answered affirmatively were classi-fied as having a farm background, and otherswere classified as having a nonfarm background.Responses were obtained from more than 99percent of sample women. Note that the term"farm" was not defined for respondents, sothere may be considerable variation in the child-hood experience of those who reported a farmbackground.

Education.-Education is classified accordingto the highest grade or year of regular schoolor college that was completed. Determinationof the highest year of regular school or collegecompleted by the respondent is based on re-

sponses to a series of questions concerning (a)the last grade or year of school attended, (b)whether or not that grade was completed, (c)whether any other schooling of a vocational orgenerally nonacademic type was obtained, and(d) whether or not such other schooling was in-cluded in the years of regular school or collegereported in (a). Information on education wasreported almost completely. Only about 1 per-cent of the data was imputed.

Occupation.-Occupation was determined byasking women: "What (is/was) your (main) oc-cupation? That is, what (is/was) your job called?What (are/were) your most important activities

37

or duties? What kind of place (do/did) you workfor? What do they make or do?" The answers tothose questions were recorded verbatim andused by coders to find the most appropriatestandard job title in the 1970 U.S. Census occu-pation classification. If the responses indicatedmore than one occupation, the primary occupa-tion was coded. If none was primary, the firstmentioned occupation was coded. Although theclassification used was very detailed, occupa-tions have been grouped into major categoriesfor this report according to the practice of theU5. Bureau of the Census. For a more detaileddiscussion, see the Department of Commercepublication, "1970 Census of Population, De-tailed Characteristics, U.S. Summary," PC (1)Dl, pp.A17-A18.

Poverty level.-The Federal Governmentperiodically estimates the amount of incomeneeded by male- and female-headed families ofdifferent sizes to purchase essential goods andservices, the "poverty threshold." (See U.S.Bureau of the Census, Current PopulationReports, Series P-60, No. 98, "Characteristics of-the Low Income Population," Table A-3.) Pov-erty level income for a family is the ratio of itstotal family income to the official poverty thres-hold for a family of its size and sex of head, ex-pressed as a percentage of the poverty threshold.

Geographic region.-Da.ta. are classified byregion of residence into the four major Censusregions: Northeast, North Central, South, andWest. Sample size varies greatly and restricts thepossibility of meaningful analyses by socialcharacteristics among smaller geographic divi-sions. The States comprising these four majorgeographic regions are:

Northeast North Central

OhioIndianaIllinoisMichiganWisconsin

MaineNew HampshireVermontMassachusettsRhode Island

North Central-Con.Northeast-Con.

ConnecticutNew YorkNewJerseyPennsylvania

MinnesotaIowaMissouriNorth DakotaSouth DakotaNebraskaKansas

WestSouth

DelawareMarylandDistrict of ColumbiaVirginiaWest VirginiaNorth CarolinaSouth CarolinaGeorgiaFloridaKentuckyTennesseeAlabamaMississippiArkansasLouisianaOklahomaTexas

MontanaIdahoWyomingColoradoNew MexicoArizonaUtahNevadaWashingtonOregonCaliforniaAlaskaHawaii

Employment.-In addition to their current

employment status, respondents were asked:". .Did you ever work for pay?" in various,specified earlier periods as appropriate. The peri-ods were: before first marriage, since first mar-

riage, between marriage and first birth, betweenfirst and second birth, between second and thirdbirth, and between the third, and last birth.Within these periods women were classified as

employed if they worked for pay and not em-

ployed if they did not. The rate of nonresponseto these questions was low, its maximum beingfor questions about employment between thebirths of the third and the last child. Employ-ment status was not ascertained for 3 percentof women asked about the period, and durationof employment was not ascertained for 9 per-cent. Missing values were imputed.

0 0 0

38

APPENDIX III

ITEMS ON NSFG QUESTIONNAIRE RELATED TO BREAST FEEDING

In study of family growth in this country, of the important things is knowinbasic faces about pregnancies and births that have.

18. Have you had baby bomyou any time?

Ye (ASK A)

Mo (SKIP TO Q. 21)

A. IF YES: Altogether, how many babies have you had born you,Including any who died very young? (Number of

live birchs)[including any who died very young? (Number of i----| Recall

live birchs)L___] Chart (C)

19i Now I’d like get information about (each of) your (baby/babtes).

(ASK A-F FOR EACH LIVE-BIRTH.)

D

A. When your (first, second, etc.) childborn? (ENTER DATE IN COL. Y OF BIRTH &

PREG. RECORD BETWEEN HEAVY LINES.)

B. What did you the baby? (ENTER INCOL. OF BIRTH & PREG. RECORD NEXT TODATE OF BIRTH.)

(ENTER C-F IN CHILD COLS. TO RIGHT)

C. Was that baby boy girl?

D. How much did (CHILD) weigh birch?

(1) Did (he/she) weigh than

5^ Ibs. less?

E. IF NOT LISTED IN HOUSEHOLD. ASK:

do have (CHILD) listed in the house-

hold. Is (he/she) still living?

IF YES. ASK:(I) Is this where (CHILD) usually lives

does (he/she)’live somewhere else

of the time?

(2) When did (CHILD) last live with youregularly?

(3) Where is (he/she) living now?

Other (SPECIFY)

IF NO. ASK:(4) When did (CHILD) die?

F. IF (CHILD) LIVED WITH MOTHER AT LEASTTWO MONTHS. ASK:(I) When (CHILD) Infant, did you

breastfeed (him/her) all?

(2) How many weeks old (he/she) when

you quit breastfeeding (him/her)altogether? RECORD VERBATIM IF RDOES NOT ANSWER IN "WEEKS."

definitely does remember

?f) Sometimes miss baby who died shortly afterbirth lived home. Have listed

all your babies now?

FIRSTCHILD

Boy Girl

Lb. Oz

More

5s lessDon’C know 8

Yes(ASK(l)).

No (ASK(4)).

Lives hereLiveswhere else.[ASK(2)&(3)1

nc:Month Year

4

^cMonth Year

Yes(ASK(2)).No

cmWeeks

98

SECONDCHILD

Boy Girl

Lb Oz

5i lessDon’C know

Yes(ASK(l)l.

No (ASK(4)).

Lives hereLiveswhere else.[ASK(2)&(3)j

11Month Year

11Month Year

Yes|ASK(2)).

No 2

cmWeeks

98

Yes

No (ASK A)

THIRDCHILD

Boy Girl

Lb Oz

5* lessDon ’.t know

YesIASK(l)).

Mo (ASK(4)). 2

Lives hereLiveswhere else.

(ASK(2)&(3)1

11Month Year

11Month Year

Yes[ASK(2)).No

cmWeeks

9798

A. IF NO; How many did miss? CO BACK TO Q. 19 AND ASK A-F FOR EACH BABYMISSED? ENTER INFORMATION ABOUT MISSED BABIESAND INDICATE PROPER BIRTH ORDER BY ARROW OM

BIRTH AND PREGNANCY RECORD AND IN COLUMNS ABOVE.

A- US. GOVERNMENT PRINTING OFFICE 1973-0-311-240 (36)

Documents

Trends in Breast Feeding Among American Mothers PUBS/A21.pdf · 2018. 10. 6. · Datafromthe NATIONALSURVEYOFFAMILYGROWTH Series 23 Number3 Trends in Breast Feeding Among American