TEEN-AGE FAMILY FORMATION

IN A RURAL COUNTY

WILLIAM D. MANGOLD
University of Arkansas

In 1975 every sixth birth was delivered by a woman under the

age of 20. For first births, that is, births signifying entry into
motherhood, one of every three was registered to a woman in
her teens. Indeed, almost one of every six new mothers in 1975
was under the age of 18 (U.S. National Center for Health
Statistics, 1976). As is well known, a major difference between
the fertility of teenagers and older women (that is, women over
the age of 20) is the preponderance of premarital conceptions
and illegitimate births among younger women. Recent studies
using varied methodologies (Whelan, 1972; Zelnick and
Kantner, 1978; U.S. Bureau of Census, 1977) document the
fact that conception following marriage is the exception rather
than the rule among contemporary adolescents. An addi-
tional characteristic of teen-age fertility is the extent that
white women generally have a substantially lower likelihood of
delivering a premaritally conceived or illegitimate birth than
is the case for most of the nonwhite population. In 1975 there
were 12.1 and 88.1 illegitimate births per 1,000 single teen-agers
among whites and nonwhites, respectively. In the same year,

AUTHOR'S NOTE: Research f o r this paper was made possible

tkrorrgli a Biomedical Support Grant. 303-3410. Actdirional slipport
was provided bj*a National histitirte of General Medical Sciences
lrairiiiig throrcgli the Ceiiter for Demographic Stitdies, Duke Utii-
versitj*,Grant Nirriiber NIGMS TO1 EM01291.
YOUTlI & SOCIETY. Vol. 10 No. 3. March 1979 257-273
0 1979 Sage Publications, Inc.
251
258 YOUTH & SOCIETY / MARCH 1979

the ratio of illegitimate t o total births among white teen-agers

was 229 per 1,000, while among nonwhites under 20 it was 747
(US.National Center for Health Statistics, 1976).
It is ironical to note that childbearing among teen-agers is
near record highs a t a time when the record low period fertility
of American women in general has been explained in terms of
“rational” forces such as implied by the “New Home Eco-
nomics” as well as the increased availability of reliable con-
traception and abortion to reduce unwanted births.
The preceding examples are indicative of two aspects of
teen-age fertility that will be examined. First, teen-age fertility,
in general, is characterized by substantial departures from the
conventional family formation process in which conception, if
it occurs, follows marriage. Second, there are considerable
differences between white and nonwhite teen-agers in propor-
tions either pregnant outside of marriage or single at delivery.
The research examines three categories of family formation,
each based on the timing of conception relative to marriage.
The fertility paths leading to each of the family formation
categories can be represented by a flow chart.
As implied by the flow chart of family formation paths, it is
necessary to examine two factors: ( I ) variables related to the
likelihood of conception and (2) variables that structure the
likelihood of marriage. Variables in the first set include coital
activity and contraception, while variables in the second set
focus on various measures of rapidity of the involvement
between the couple prior to conception. I believe that the
two sets of variables are not independent of each other, but are
interdependent in a manner such that the values of one set
structure the other. More specifically, variables having the
greatest impact on the likelihood of conception-coital
frequency and contraceptive use-are at the same time affected
by the relationship between the couple. The relationship
between the couple, or dating relationship, as it will be called, is
viewed as a major determinant of the likelihood that a single
pregnant woman will be married before delivery.
 Mangold / TEEN-AGE FAMILY FORMATION 259

Concept Ion c f t c i

\
!'arrinpe h e f o r e

Figure 1

Studies of marital fertility have illustrated the importance

of communication as a factor in reducing unwanted births,
primarily through more effective contraception (Rainwater,
1960, 1965; Hill et al., 1959). Among single teen-agers, a
considerable amount of premarital sexual activity takes place
between couples in a stable, on-going relationship, and for
many as a prologue for marriage. To the extent that the rela-
tionship between the couple develops over a longer period of
time, the chances of shared expectations relative to the role
of sexual activity and fertility desires should be greater. The
effect of the longer period of heterosexual involvement should
increase both coital activity and contraceptive use.
The length of the dating relationship should be related t o
the type of contraceptive used. Two arguments can be given to
support this assumption. First, regularity in sexual activity is
associated with routinized contraceptive use among teen-agers
(Zelnick and Kantner, 1973, 1977). Second, couples in a more
stable relationship-especically very young women-are
presented with a lessened "pseudo-moral" barrier which pro-
scribes sex in the absence of love. The pill, unlike many other
methods, is not situation specific but requires planning in
260 YOUTH & SOCIETY I MARCH 1979

advance of sexual relations. Such planning would be easier in

a stable relationship of longer duration.
Given that a single woman becomes pregnant, her chances
for marriage are also shaped by the dating relationship. To the
extent that the dating relationship has lasted a longer period
of time, then pregnancy, if it occurs, will more likely be fol-
lowed by marriage. Several reasons can be given in support of
this contention. First, there is a self-selective process that
modifies the pool of potential husbands. Young men whose
primary interests are sexual are removed as the dating interval
increases without sexual access. Relationships developing over
a longer period of time are more conducive to the development
of shared expectations-especially the role that sexual rela-
tions plays between the couple. Finally, relationships devel-
oping over a longer period of time are more likely to be
embedded in a wider set of expectations from parents t o peers.

METHOD

The data on which the research is based consists of inter-

views with 114 young women in a rural county of North Caro-
lina. The sampling frame consists of resident females of the
target county who had their first live birth prior to their
twentieth birthday. Eligibility was determined by an inspection
of all birth records for the target and contiguous counties and
yielded a total of 560 births to teen-age women. From informa-
tion contained on the birth records, about one-half of the total
births were delivered to married women. By race, 36% of black
and 88% of white births were delivered to married women.
Whites were disproportionately selected over blacks; while by
marital status, blacks were selected proportionately to their
numbers in the birth records. Among whites, unmarried
women were overselected to compensate for their low frequen-
cies in birth records. The final sample is shown in Table 1 and
'reflects a completion rate of 90% of locatable whites and 94%
 Mangold / TEEN-AGE FAMILY FORMATION 261

TABLE 1
Distribution of Women by Marital Status at Delivery
for Interviews and Birth Records by Race

li’hite Rlack
Marital Status Sample Records Sample Q.ecords

Single

of blacks. A complete description of the sample can be found

in Mangold (1974).

RESULTS

Two of the more frequently encountered explanations of

premarital pregnancy and illegitimacy is that there is a lack of,
or a reduced commitment to, a “norm of legitimacy” and the
lessened stigma attached to such births accounts for their
differential occurrence (Myrdal, 1944; Christensen, 1960;
Pope, 1969; Zelnick and Kantzer, 197 1). Others, such as Goode
(1960), argue that commitment to a norm is not the same as
compliance, while various studies have found that illegitimate
birth is typically unwanted (Gibbs et al., 1973; Furstenburg,
1971; Presser, 1974; Zelnick and Kantner, 1974, 1978).
Among the young women in the study, slightly less than one-
third (31) claimed that they were trying to become pregnant; ,
while less than one-half (54) reported that they “wished” for
pregnancy. As can be seen in Table 2, there are marked varia-
tions in the response to these two questions between black and
white teen-agers as well as among respondents in each of the
 262 YOUTH & SOCIETY / MARCH 1979

TABLE 2
Indicators of the Extent that Pregancy was Desired by
Family Formation Categories and Racea

h3ite Black

Indicator IIC PIlC UB Total PNCb UB Total

A l l Komen 100%
(17)

h'ished for 23.5

Pregnancy (4 1

Tried t o
Become 58.8 38.9 11.8
Pregnant (20) (7) (2)

Negative
Reaction by 3.8 44.4 52.9
Mother (3) ( 8) 19)

Negative
Reaction by 11.1 17.6
Partner (2) (3)

a. Codes for family categories: M C = Marital Conception PMC = Premarital

UE = Illegitimate Birth.
b . Throughout t h e paper t h e t w o maritally pregnant black w o m e n have been
combined with premaritally pregnant.

family categories. I n general, white teen-agers expressed greater

agreement than did the young black women, and maritally
pregnant women report a higher number of positive answers to
the questions than do women in other categories. Premaritally
pregnant women report a higher number of positive responses
than do the women with an illegitimate birth. None of the 14
unmarried white and two of the 31 unmarried black mothers
reported that they actually tried to become pregnant.
Since it could be argued that the recall of pregnancy inten-
tions could be dependent upon subsequent reproductive
events, another series of indicators were used. When the
subjects were asked to report the initial reaction of their
mother and partner, a pattern that is consistent with the earlier
responses emerges. Again white teen-agers report a more
favorable (e.g., less negative) response by their parents and
 Mangold / TEEN-AGE FAMILY FORMATION 263

boyfriend t o the pregnancy. A systematic pattern emerges in

that the maritally pregnant have the lowest number of negative
answers, premaritally pregnant have a higher number, and the
mothers with an illegitimate birth have the highest number of
negative responses.
As the large majority of young women finding themselves
pregnant prior to marriage typically did not want t o become
mothers at the time, explanations of their behavior must be
found elsewhere. In this section attention is directed toward
an examination of differences with respect t o the “conception
variables” of coital activity and contraception.
Table 3 shows the distribution of women by coital frequency
for each of the family categories. The data shown are the
responses to a question about coital frequency from the initia-
tion of sexual activities until the knowledge of the pregnancy.
Among white teen-agers the median frequency reported was
10.0 per month, while among blacks it was 7.0. Although it is
difficult to compare these absolute frequencies with previous
studies due to differences in the population and nature of the
questions, the differences between white and black women are
consistent with the findings of Zelnick and Kantner (1972,
1977).
Examination of the detailed distribution of the coital
frequencies for women in each of the family categories reveals
that, regardless of race, women single a t delivery report the
lowest frequencies. Women reporting the highest frequencies
are the white teen-agers with a maritally conceived birth.
Notice that a distinction is made between women with and
without premarital coitus. Given that marriage affords greater
opportunity for sexual activity, it is understandable that
women becoming sexually active following marriage report the
highest frequencies.
The distribution of women reporting contraceptive use prior
to the first birth is shown in Table 4. Again a distinct-ion is
made between women with and without premarital coitus. By
race, 47% of white and 35% of black women report using some
 TABLE 3
Distribution of Women by Monthly Coital Frequency Prior to Conception by
Family Formation Categories, Premarital Coitus, and Race

h'hite Black
?lC PMC IJ B Total PF!C (lI3 Total

Premarital C o i t u s
Frequency
Per >tonth No Yes

A l l women 100% 100% 100% 100%

(161 (181 (181 (14)
16+ 43.7 16.7 21.4
(71 (31 (3)

5 - 15 56.3 38.9 28.6 64.1

(9) (71 (41 (111

llnder 5 16.7 44.8 SO.O 27.3 29.4 41.7

(31 (8) (71 (18) (V w-l?
Median 14.0 15.0 7.3 6.0 10.0 8.5 6.2 7.0
 TABLE 4
Distribution of Women by Use and Type of Contraception Prior to First Birth by
Family Formation Categories, Premarital Coitus and Race

White Black
Contraceptive NC PFIC UE Total PMC UB Total
Use
Premarital Coitus
No Yes

Method

Pill- 65.3 27.8 5.6 7.1 24.2 0.n 0.0 0.0

Iun (9) (51 (1) (11 (16) (0) (0) (0)

Condom 6.3 5.6 22.4 21.4 13.6 17.6 35.5 29.2

(2) (1) (41 (3) (91 (31 (111 (141
Othersa 12.5 25.0 5.6 7.1 9.1 17.6 0.0 6.3
(2) (2) (11 (1) (6) (31 (0) (3)
N
8 a. Includes foam, jelly, douche, withdrawal, and dlaphram.
266 YOUTH & SOCIETY I MARCH 1979

form of contraception prior to their first pregnancy. The group

differing most are the maritally pregnant whites initiating
sexual activity following marriage. Of these women, 75%
reported use which is substantially higher than was true
of women in any of the other categories. As can be seen, there
is little difference between the proportions reporting con-
traceptive use among women in other family categories. It is
significant to observe that the most important apparent aspect
of contraceptive use was whether sexual activity occurred
before or after marriage.
When the type of method is examined, women with a mari-
tally conceived birth stand out clearly-all but two women
reporting use of the pill were in this category. Apparently
there is a much greater reliance on the male as the initiator
of contraception among the unmarried as reflected by the rela-
tively large proportion reporting use of the condom in this
category.
Thus far we have examined differences between young
women in each of the family categories for the extent that
pregnancy was desired and the conception variables of coital
activity and contraceptive use. Next, attention is directed
toward an analysis of the tempo of the heterosexual relation-
ship between the couple prior to pregnancy. Three indicators
of the tempo of the heterosexual relation consisted of the
amount of time between the couple's first date and the decision
to date on a regular or steady basis, the interval between dating
on a regular basis and the initiation of coitus, and the amount
of time between the initiation of sexual relations and knowl-
edge of pregnancy.
The median interval in weeks for each of the tempo indi-
cators is displayed in Table 5. Considered either separately
or as a set, a consistent pattern is evident: mother; of an
illegitimate child reported the shortest period of involvement
with the baby's father and the maritally pregnant reported the
longest intervals. The only exception to this pattern is the
identical 23.0 weeks reported by the premaritally pregnant and
 hlangold / TEEN-AGE FAhllLY FORhlATIOS 267

TABLE 5
Length of Time Between the First Date and Dating on a Regular
Basis, Initiation of Coitus, and Knowledge of Pregnancy by
Family Formation Categories and Race

Median
Interval FlC
____---
hhite
PhlC un -
TOTAL n:c
Black
UB TOTAL
(Iceeks)

F i r s t Date
Until 4.7 2.2 1.5 3.3 30.5 5.1a 17.3
Rcmlar (34) (15) (14) (66) (17) (3C) (471
Date

Regular
Date 3.Ob 40.0 15.5 29.2 21.0 13.7' 16.3
Until (18) (18) (14) (50) (17) (50) (47)
CoitGs

Coitus
Lbtil 52.4 2b.O 16.0 37.5 23.0 23.0 23.0
Prcgnancy (34) (If) (14) (66) (17) (31) (48)

a. Excludes one woman who claimed that she never dated the baby's father o n
a regular basis.
b. Restricted t o women with premarital coitus.
c. Excludes one woman who refused t o answer.

the unmarried black teen-agers. Analysis of the black-white

differentials reveals that the black teen-agers took longer to
date on a regular basis but following the commitment to date
on a regular basis were less able to delay the initiation of coitus.
Another way of summarizing the tempo differences among
women in each of the family categories is to include informa-
tion on the level, as well as the length, of sexual involvement.
A relative conception index was computed based upon the
length of time the woman and partner were sexually active
and the reported level of activity. The values for the index are
shown in Table 6. Two striking features of the values shown are
the large differences among women in each of the family
categories. Women with a n illegitimate birth obtain a relative
risk of pregnancy that is 4.16 times greater than obsei-ved
among the maritally pregnant. For the premaritally pregnant
the relative risk is 2.60 times greater. As can be seen, the dif-
268 YOUTH & SOCIETY / hlARCH 1979

TABLE 6
Relative Conception Index by Family Formation Category and Race

Family R e l a t i v e Conception Jndex"

Category \it+t e Black lotal

Total

Mar it a1 1.00 1.00

Pregnancy (34) (34)

Legitimated
Pregnancy

U nl egi t i m zt ed
Pregnancy

a. Values s how n represent t h e ratio of t h e central pregnancy rates f o r each cate-

gory t o t h e central pregnancy rate for white females with a marital pregnancy.
T h e num ber in each c a te gory i s s hown in parenthesis. T h e central pregnancy
rate = l / ( e x 1) w he re e = e xpos ure is t h e interval in w e e k s f r o m first c o i t u s until
learning of pregnancy, a n d f = coital frequency per week.

ference between black and white teen-agers is relatively small.

In spite of the restricted age range in the sample, blacks were
considerably younger than whites. An analysis of the age
distributions of the two groups revealed that whites were, on
the average, about 1.4 years older at the time of their first
birth. The age difference between whites and blacks a t the
time of delivery of a given birth order is one of the most
consistently documented aspects of black-white fertility
differences. There are two major reasons for this: a higher
proportion of blacks are sexually active at the younger ages
(Zelnick and Kantner, 1974) and black women are less likely to
use contraceptives prior to the first birth (Rindfuss and
Westoff, 1974). From data in this study it is possible to
examine the extent to which age differentials between black
and white teen-agers are attributable to an earlier initiation
of sexual activity or differences in contraceptive use.
Table 7 presents the mean age of women at their first birth
and the age when sexual activity began. As can be seen, the
contraceptors were about 1.4 years older at their first birth
 Mangold / TEEN-AGE FAhllLY FORhlATiOS 269

TABLE 7
Age at First Live Birth and First Coitus by
Contraceptive Use and Race
kkite Elack Total

Interval Coritraceptivc
Use

Hean Age at A l l h'omen 17.4 16.3 17.0

First Coitus (66) (48) (114)
Users 17.7 17.1 17.5
(31) (17) (48)

Xon-users 17.2 15.9 16.6

(35) (31) (66)

Bean Age a t A11 I:'or.cn 18.7 17.3 18.1

First Eirth (66) (48) (114)
Users 17.7 17.1 17.5
(31) (17) (48)
Eon-users 18.1 16.9 17.5
(35) (31) (66)

than were the nonusers (18.7 versus 17.3). The youngest are the
black nonusers, who averaged 16.9 years of age.
Analysis of the age at the onset of sexual activity reveals
that the major source of difference in age at first birth between
black and white teen-agers is primarily due to a divergence in
patterns of contraceptive use between black and white women
and, more specifically, a relationship between age and con-
traceptive use among black teen-agers. While there is little
difference in the age of white users and nonusers (roughly 0.5
years), there is a 1.2-year difference between black users and
nonusers. The average age a t first intercourse among black
useri is 17. I , or 0.3 years younger than white women. Among
black nonusers there is a much larger difference, the greatest
disparity being 1.8 years between white users and black non-
users.
270 YOUTH & SOCIETY / hlARCZ1 1979

It would appear that the difference in age at first birth is

attributable to a lower proportion of black teen-agers using
contraceptives a t all ages and younger blacks being less likely
to use contraceptives. Early entry into sexual activity among
the black teen-agers is not accompanied by contraception and
the resulting first-birth interval is shorter.

DISCUSSION

From the preceding review of differences between black

and white teen-agers and women in each of the family forrna-
tion categories, it is possible to piece together some of the
significant aspects of family formation among women in the
study. Unmarried women had a higher risk of pregnancy on an
exposure basis than did women in other categories. These
women also had known the baby's father for a shorter time
than did other women. It is suggested that the dating relation-
ship is a significant element in both the likelihood of preg-
nancy, and, following pregnancy, the probability of marriage.
Unmarried women were more likely to become pregnant within
a shorter period of time, had dated the baby's father for a
shorter period of time before becoming pregnant, and, there-
fore, had a shorter period of time for the development of shared
expectations relative to the role of sex, childbearing, and
marriage.
In addition to the relationships involving the interval until
the initiation of sexual relations as an element in the chances of
a single, pregnant woman marrying at delivery, there are also
implications for family structure that have particular signifi-
cance for the black-white illegitimacy difference. Black teen-
, agers tended to enter into sexual relations not only'more
rapidly than did whites but they did so a t a younger age. As
there was a substantial difference in contraceptive use by black
women at the younger ages, early entry into sexual activity
meant that the conception delay was reduced. Thus, while
 hlangold / TEEN-AGE FAhlILY FORMATION 271

there was little difference in the ages of white and black women
among contraceptive users, the black nonusers were consider-
ably younger than their white counterparts. The fact that the
black teen-agers became pregnant at a younger age implied a
more youthful father of the baby. The net effect of this age
factor is t o further depress the chances of marriage following
pregnancy as the accumulation of resources required for
marriage tend to vary directly with age.
Finally, one additional facet of black-white differences
requires discussion. The higher prevalence of pregnancy
outside of marriage among black teens cannot be regarded as
either a greater desire for childbearing or a reduced commit-
ment for marriage. As indicated by a number of separate items,
black women were more hesitant about becoming pregnant
than were whites. Black teen-agers were less likely to report
that they had wished for, or thought that others wanted them
to have, a baby. They were almost universal in their denials
that they had tried to become pregnant. Within comparable
family categories-the unmarried or the premaritally preg-
nant-blacks reported substantially lower pregnancy desires
for motherhood.
In this respect, motherhood among blacks in the study
presents an enigma: while black teens denied either intent or
desire to become a mother, there was little conscious inter-
vention to avoid pregnancy in spite of an apparent ability to d o
so. One explanation of this contradiction may be found in the
nature of the black family and the relative lack of alternatives.
Over 40y0 of the black women in the study were raised in a
female-headed family. Family size among the mothers of the
black women in the study tended to be quite large, averaging
7.5 children per family. Reflecting the rural, impoverished
nature of the county, employment opportunities were limited.
The primary source of employment for black women in the
study was that of a farm laborer or as a domestic worker.
Nearly 75% of the black teen-agers in the study had never held
a regular or a full-time job.
 272 YOUTH & SOCIETY I hfARCH 1979

For black adolescents in the study, motherhood is a way of

life-there is little conscious intent to become a mother and
there is little intervention since both intent and intervention
imply the availability of alternatives. The personal ramifica-
tions of motherhood become evident only in the face of viable
alternatives, and these alternatives are not available for many
of the young black women in the study. Thus, while there may
have been few rewards for having a baby, there also were few
penalties.

REFERENCES

BOWERhlAN, C., et al. (1966) Unwed Motherhood: Personal and Social Con-
sequences. Chapel Hill: Institute for Research in Social Science, University of
North Carolina.
CHRISTENSEN, H. (1960) ”Cultural relativism and premarital sex norms.” Ameri-
can Sociological Review 25: 30-39.
FURSTENBURG, F. (I97 I) “Birth control experience among pregnant adolescents:
The process o f unplanned parenthood.” Social Problems 19: 192-203.
GIBBS, Jr., et al. (1973) Unwed hlotherhood and Premarital Pregnancy: Practical
Significance and Sociological Explanation. Final Report to the National Institute
of Child Health and Human Development.
GOODE, W. (1960) “Illegitimacy in the Caribbean social structure.” American
Sociological Review 25: 21-30.
HILL, R.. et al. (1959) The Family and Population Control: A Puerto Rican Experi-
ment in Social Change. New Haven: College and University Press.
hf ANGOLD, W.(1974) Teenage Fertility and Family Formation. Doctoral Disserta-
tion. Center for Demographic Studies. Duke University. Durham, N.C.
hlYRDAL. G. (1934) An American Dilemma. New York: Harper and Row.
POPE. H. (1969) “Negro-white differences in decisions regarding illegitimate chil-
dren.”JournaI of hlarriage and the Family 31: 756-764.
PRESSER, H. (1974) “Early motherhood: Ignorance or bliss?” Family Planning
Perspectives 6: 8-14.
RAINWATER, L. (1965) Family Design: Marital Sexuality, Family Size and Con-
traception. Chicago: Aldine.
. ___ (1960) And the Poor Get Children. Chicago: Quadrangle.
RINDFUSS, R.. and C. WESTOFF (1974) The initiation of contraception. Demo-
graphy I I: 75-88.
U.S. Bureau of the Census (1978) Trends in Childspacing. June 1975. Current Popu-
lation Reports, Series P-20. No. 315. Washington. D.C.: U.S. Government
Printing Office.
 Mangold / TEEN-AGE FAhllLY F O R h l A T l O S 273

U.S. National Center for Health Statistics (1976) Advance Report: Final Natality
Statistics, 1975, Vol. 25, No. 10, Supplement. Washington, D.C.: U.S. Govern-
ment Printing Office.
WHELAN, E. (1972) "The temporal relationship of marriage, conception, and birth
in hlassachusetts." Demography 9: 399-414.
ZELNICK, hl., and J. KANTNER (1978)"First pregnancies to nomen aged 15-19:
1976 and 1971." Family Planning Perspectkes I0 11-20.
(1977)"Sexual and contraceptiveexperience ofyoung unmarried women in the
United States, 1976 and 1971." Family Planning Perspectiyes 9: 55.
---(1974) "The resolution of teenage first pregnancies." Family Planning Perspec-
tives 6: 83.
_-- (1973) "Contraception and pregnancy: Experience of young unmarried a o m e n
in the United States." Family Planning Perspectives 5: 21-35.
__- (1972) "Sexual experience of young unmarried women in the United States."
Family Planning Perspectives 4 9-18.
--- (1971) "Exploratory studies of Negro family formation-factors relating to
illegitimacy." Studies in Family Planning 60: 5-9.

IViIliatii D. htati~oliiis preserit!r an Assistant Profenor of Svriology ut the

Universiir of Arkansas. He received his PI].D. from Duke Utiiversit.v in 1974
where he specialized in Social Deniograpiij,. His riiajor research interem are
iri ihe urea of atiolescetir fertility and fami!r siriirtiire.