Académique Documents
Professionnel Documents
Culture Documents
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
National Council on Family Relations is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marriage and Family.
http://www.jstor.org
LARRYL. BUMPASS
ANDJAMES
A. SWEET University of Wisconsin-Madison
ANDREWCHERLIN Johns Hopkins University*
913
914
married,remainednearlyconstant.Thatis to say,
the fall in the proportionever marriedamong
young adultswas compensatedby the increasein
the proportionwho had evercohabited.The first
experienceof livingin a union was not occurring
at a laterage, on average,althoughthe firstunion
increasinglywas a nonmaritalone. After about
1981 or 1982, however, the situation changed.
From that point, the proportionever marriedfell
morerapidlyand wasnot fullycompensatedby an
increasein cohabitation,so that the proportion
experiencinga first union by a givenage declined
substantially.Similarresults have recentlybeen
reported for Sweden (Hoem and Rennermalm,
1982),Australia(Bracherand Santow, 1988),and
Canada(Rao, 1990).
Until recently there have not been adequate
datafor an assessmentof the relationshipbetween
changingmarriageand cohabitationin the United
States. Fortunately,we are now able to do this
with the recentlycompletedNational Surveyof
Familiesand Households(NSFH).
THENATIONAL
SURVEYOF
FAMILIES
ANDHOUSEHOLDS
The NSFH is a national sample survey of 13,017
respondents, conducted in 1987-88. In addition to
a main sample of 9,643 persons aged 19 and over,
we oversampled certain population subgroups.
The oversample included households containing
single-parent families, stepfamilies, recently married couples, cohabiting couples, blacks, Chicanos, or Puerto Ricans. In each selected household, a randomly selected adult was interviewed.
A self-administered form was also filled out by
parent in a child's household). Interviewsaveraged about one hour and forty minutes in length.
The NSFH was designed to provide detailed in-
915
BEFORE AGES
20 AND 25:
MALES
80-
60
C
c
40
LEGEND
Age 20 EverMarried
"70
20 EverMarried
or Cohabited
----Age
20---C
a)
Age 25 EverMarried
Age 25 EverMarried
or Cohabited
0
"
1945
1955
1950
1960
1965
BirthCohort
BEFORE AGES
20
100
--
80 --
14
60O
LEGEND
a
-
Age 20 EverMarried
Age 20 EverMarried
Cohabited
co
20or
c
Age 25 EverMarried
CL
1945
1950
1955
1960
BirthCohort
1965
Age 25 EverMarried
or Cohabited
916
TABLE 1. RELATIVE CHANGE IN THE PROPORTIONS EVER MARRIED AND EVER IN UNION BY
Category
Percentage
Ever Married
1970 1985 70 Change
Percentage
Ever in Union
1970 1985 % Change
29%
21
38
30
28
23%
13
33
25
15
-20%
-38
-11
-15
-53
47
44
16
49
28
10
4
-37
-39
75% 69%
59
69
78
82
71
78
66
61
78
82
69
76
74
62
0 Change Married
-% Change Unioned
07 Change Married
59%
47
71
67
29
114
38
37
-8%
-14
-4
-9
-7
67%
59
76
61
83
-3
-10
-10
84
57
63
917
TABLE 2. RELATIVE CHANGE IN THE NUMBER OF YEARS IN MARRIAGE, AND NUMBER OF YEARS IN A
UNION, BEFORE AGE 25: COHORTS REACHING THESE AGES AROUND 1970 AND 1985
Category
Total
Males
Females
Whites
Blacks
Education
0-11 years
12 years
College
3.08
2.52
1.42
-21
-24
-37
educated.
About 40% of cohabiting unions in the United
States break up without the couple getting married, and this tends to occur rather quickly. By
about one and one-half years, half of cohabiting
couples have either married or broken up (Bumpass and Sweet, 1989a). Consequently, the number of years in a union before age 25 may have declined even when the experience of a union by that
age has remained relatively stable. Table 2 shows
the number of years in marriage and number of
years in any union before age 25. There was a
29% decline in the number of years in marriage
before age 25, compared to the 15% decline in the
number of years ever in a union by this age. Part
of this, of course, reflects the decline in years married by age 25 as a consequence of later marriage
among those who married before this age, and
part also reflects the decreased stability of marital
unions. However, it is also the case that cohabitation offset the decline in years in union less than
the decline in the proportion ever in union by age
25 (48% as compared to 67%). Nonetheless, the
increase in young adult years spent "single" is only about half that suggested by declining marriage
rates.
Though the decline in number of years married
before age 25 was greater among blacks, there was
an identical 17% decline for both blacks and
whites in the number of years in a union before
that age.
In both Tables 1 and 2, it is clear that cohabitation has compensated for declining marriage least
among persons who have attended college. For
example, it offsets 84% of the decline in marriages before age 25 among persons not completing
high school, compared to 63% of this decline
4.32
3.33
2.34
3.87
3.11
1.86
-10
-07
-21
-% Change Union/
% Change Married
48%
47
43
43
61
52
71
43
among those who attended college. This is contrary to the image of cohabitation as a college student phenomenon. The popular treatment of cohabitation has focused on college graduates or
college students in urban areas-rather than on
the consensual unions that were more common
among the lower class. Thus there was the impression that cohabitation was an innovation of college students in the 1960s, which then spread during the 1970s and 1980s to the rest of the population. Taking into account "common-law" marriages in the lower class, we might expect that cohabitation was rare outside the lower class before
it was adopted by well-educated young adults in
large metropolitan areas and then diffused widely
across the middle and working classes. Evidence
relating to this model has been slim. One Swedish
study reports, to the contrary, that rates of consensual union formation were consistently higher
among young working-class women than among
young women from the bourgeoisie across birth
cohorts, and consistently higher among nonstudents than among students (Hoem, 1986). The
trends by education in Figure 3 speak directly to
this issue. In order to control for possibly confounding compositional changes over cohorts, the
data graphed in this figure have been adjusted for
age at marriage and race.
This figure does not support the diffusion
model. Rather, it indicates that among the birth
cohorts of the 1930s-who reached their midtwenties in the late 1950s-cohabitation in young
adulthood was restricted to a small minority in the
lower educational groups. And it is among these
same groups that the rise in cohabitation began in
the late 1950s. By the 1960s, when the birth cohorts of the 1940s entered adulthood, the rise was
918
60
50
40
30
0
0
20
-- LEGEND
/0-11
10/
YEARS
12 YEARS
COL4+
1950
1955
COL1-3
0-
o
1930
1935
. -
...
1940
1945
1960
Midpointof 5 YearBirthCohort
PostmaritalUnionFormation
Because of our high levels of marital instability,
large proportions of the population experience
separation or divorce. At the same time, rates of
remarriage have fallen rapidly. We have seen that
cohabitation has offset much of the decline in first
marriage rates; to what extent has it also compensated for declining rates of remarriage?
We know that cohabitation is even more common among separated and divorced persons than
among the never married. Sixty percent of persons
who remarriedbetween 1980 and 1987 lived with a
partner before the remarriage-46% only with the
person they then married and 14% with someone
else (Bumpass and Sweet, 1989a). Table 3 presents
the proportion remarrying, and the proportion
forming any union, within five years of separation
for cohorts that separated from 1963 to 1967 and
from 1977 to 1981 (cohorts that had been separated for five years around 1979 and 1984, respectively). Although the proportion marrying
within five years of separation declined 16% between 1970 and the early 1980s, the proportion
who had formed a new union within five years actually increased slightly. Hence, cohabitation has
compensated fully for the fall in remarriage.
There has been no decline in the pace of union
formation following marital dissolution, although
919
Percentage Remarried
within Five Years
1970a 1984b %7o
Change
Category
Total
Males
Females
49%
54
46
42%
45
38
Percentage in Union
within Five Years
1970a 1984b 07 Change
-16%
-17
-16
58
67
52
62
70
56
7%
4
8
riage rates to be lower among the previously married because of their older age distribution.
Whether there is a further disinclination to marry
associated with having been previously married is
a topic that warrants further examination, and
one on which we will see more evidence subsequently.
The common image of cohabiting couples as
college students, or at least young couples, does
not usually include a family with children. Yet, 4
of every 10 such couples have children present
(Table 5). This proportion is one-third among the
Children
No children
Couple's
One partner's
Total
Yes
Ever Married
No
53%
7
39
100
65 %o
16
22
100
Total
60%
12
27
100
Duration
Less than 1 year
1-1.9 years
2-2.9 years
3-3.9 years
4-4.9 years
5 + years
Total
Yes
30%
17
12
9
9
23
100
Ever Married
No
39%
19
11
8
5
17
100
Total
36%o
18
11
8
7
20
100
never-married and almost half among the previously married. One-sixth of never-married cohabiting couples have a child that was born since
they began living together. As we noted in an earlier analysis of children's single-parent experience,
this represents a significant component of unmarried births (about a quarter) that are not born into
single-parent households (Bumpass and Sweet,
1989a).
Further, the children in cohabiting households
are not all young children. In Table 6 we see that
one-quarter of the households with children have
children age 10 or older, mostly living with previously married parents. In thinking about the
meaning of cohabitation and the dynamics of
cohabiting households, it is critical to keep in
920
mind that issues of parentingand stepparenting late to differencesand similaritiesbetween coare very much a part of the picture.
habitationandmarriage.One suchsequencewasa
series of questions concerning how important
various factors are in thinkingabout whetheror
Attitudes toward Cohabitation
not to live with someone withoutbeing married.
We turn now to severalattitudinalitems thatre- Responseswere on a 7-point scale rangingfrom
"not at all important" to "very important."
7 shows the percentageof cohabitorswho
Table
TABLE 6. AGE OF YOUNGEST CHILD IN HOUSEHOLD OF
COHABITATING RESPONDENTS, BY WHETHER
respondedin the two categoriesat each end of the
RESPONDENT WAS EVER MARRIED
scale.
Ever Married
The only item that was regardedas "imporYes
No
Total
Children
tant" by a majority of the respondents was
54%
64%
No children
"couples can be sure they are compatiblebefore
60%0
21
16
24
0-4
A quarterof the respondentsreported
marriage."
7
10
14
5-9
that
opportunityto "share living expenses"was
10
16
5
10-17
100
100
100
Total
important.
TABLE
7.
Reasons why a person might WANT to live with someone of the opposite sex without being married.
How important is each reason to YOU?
Not Important
Important
Male
Female
Male
Female
Response
It requires less personal commitment than marriage.
46%
48%
18%
14%
It is more sexually satisfying than dating.
17
18
49
59
It makes it possible to share living expenses.
26
28
32
29
It requires less sexual faithfulness than marriage.
10
12
64
69
51
56
18
16
Couples can be sure they are compatible before marriage.
It allows each partner to be more independent than marriage.
17
19
36
41
Reasons why a person might NOT want to live with someone of the opposite sex without being married.
How important is each reason to YOU?
Not Important
Important
Male
Female
Male
Female
Response
a.
b.
c.
d.
e.
f.
a.
b.
c.
d.
e.
f.
g.
It is emotionally risky.
My friends disapprove.
My parents disapprove.
It is morally wrong.
It is financially risky.
It requires more personal commitment than dating.
It requires more sexual faithfulness than dating.
13%
4
8
6
7
19
24
18%
4
11
9
7
25
28
47%
84
71
75
64
43
42
44%
82
61
60
55
40
43
Better
Same
Worse
Male Female
Male Female
Male Female
Aspect of Life
a. Standard of living
19
18
74
76
7
6
b. Economic security
24
32
67
61
9
7
c. Overall happiness
30
36
57
57
13
7
d. Freedom to do what you want
11
9
59
74
30
17
Economic
e.
10
75
14
78
11
12
independence
f. Sex life
22
14
68
81
10
5
with
others
14
12
73
80
13
6
g. Friendships
h Relations with parents
22
24
71
72
7
4
28
38
i Emotional security
63
57
9
5
Note: Cohabiting respondents age 35 and younger were asked, "How do you think your life might be different if
you were married now?" A 5-point scale was used for responses. The "better" category above includes "somewhat
better" and "much better." "Worse" includes "somewhat worse" and "much worse."
921
Perceived Stability
Marriage Plans
Definite plans
Think marry
Marry someone
Not marry anyone
Total
Yes
40%
21
6
30
100
Ever Married
No
50%
31
6
13
100
Total
47%
27
6
20
100
922
Troublea
Variable
Duration
0
1-2.9 years
3+ years
Education
Less than 12 years
12 years
College 1-3 years
College 4+ years
Sex
Males
Females
Race/ethnicity
Blacks
Whites
Mexican Americans
Presence of children
No children
R's and partner's
R's, not partner's
Partner's only
Income
< $10,000
$10,000-19,999
$20,000-29,999
$30,000-39,999
$40,000 +
Age
Under 25
25-29
30-34
35+
Married before
Yes
No
Total
Unadjusted
Adjusted
Marry Partnerb
Unadjusted Adjusted
Never Marryc
Unadjusted Adjusted
43%
58
48
39%
54
54
82%
78
61
82%
76
62
13%
16
32
14%
16
28
37
54
54
48
45
51
52
43
66
76
75
77
73
72
72
78
28
19
19
14
22
22
21
13
42
56
43
54
73
73
76
70
21
20
17
24
40
53
28
44
52
36
75
74
65
74
74
67
21
18
31
22
19
28
46
62
52
49
47
62
48
56
71
76
81
62
68
79
89
62
22
24
13
32
25
20
7
31
41
45
44
56
59
40
44
42
54
62
73
75
70
76
79
66
67
70
77
86
19
19
24
19
16
26
25
23
18
10
56
54
54
33
58
52
51
35
83
80
58
67
79
76
62
72
9
16
38
25
12
20
36
21
42
54
47
50
63
81
64
80
31
13
27
16
49%
73%
20%
aProportion responding "Yes" to "During the past year, have you ever thought that your relationship might be in
trouble?"
bProportion responding "Yes" to either "Do you and your partner have any definite plans to get married?" or
"Do you think that you will eventually marry him/her?"
CProportionresponding "No" to "Do you think that you will eventually marry someone else?" That question was
asked of persons responding "No" to the two preceding questions.
CohabitationandDecliningRatesof Marriage
marriage, compared with 73% among all primary
respondents). Among the couples for which we
have joint data, there is a fairly high level of
agreement about marriage expectations. Sixtynine percent agree that they either plan or expect
marriage, and 13% agree that they do not expect
to marry each other. At the same time, this means
that there is disagreement over whether marriage
is expected in about one-fifth of the couples in
which at least one partner expects marriage.
Hence, the relatively high instability of cohabiting
unions is not surprising when we consider that
about a third either disagree about marriage or do
not expect marriage.
As we might expect, cohabitors are not unaware of the potential shakiness of their relationship. Almost half say that they have thought that
their relationship might be in trouble in the last
year-and in three of every four cohabiting relationships, at least one partner reports having
thought the relationship was in trouble. Clearly,
there is a good deal of uncertainty among cohabitors about the potential stability of their union.
Compared with married respondents and adjusted
for duration and age differences, cohabitors are
almost twice as likely to report that they have
thought their relationship was in trouble over the
past year.
Table 10 provides information on differences
among cohabitors with respect to whether they
have thought their relationship might be in trouble, whether they think they will marry their partner, and whether they think they will ever marry.
Looking at the first panel of Table 10, we
might expect that disruption would prune out less
well-adjusted relationships. This would lead to a
negative relationship with perceptions that the
relationship was in trouble and a positive one with
marriage plans. Relationships of over a year duration are more, rather than less, likely to agree to
the "trouble" measure, while those of three or
more years duration are less likely to expect to
marry the partner and particularly likely to say
they will never marry. Apparently the dominant
selection process with increasing duration is associated with marriages of better-adjusted couples.
In addition, some subset of longer-duration couples who plan never to marry may represent more
traditional "common-law" marriages.
There is surprisingly little variation in these
items by education, though the "trouble" measure peaks among those with some college and
923
924
other factors). The latter is somewhat surprising-perhaps parents are less likely to express
opinions about their children'smarriagechoices
after a separationor divorce.
Cohabitationis clearlya stepbetweenmarriage
and datingwithrespectto the constraintsit places
on individualfreedom. Cohabitorsfeel marriage
would makeless differencein this respectthan do
others. Net of other factors, about a quarterof
the cohabitorssay that their freedomto do what
they want would be worse if they were married,
compared to about two-fifths of persons not
cohabiting.As we observedfor this item among
cohabitingcouples in Table 8, men seem much
more concernedabout the effect of marriageon
their independence than women (49%0 vs. 31070).
The large increasesin the proportionnevermarried amongpersonsin theirearlytwentiesis commonly interpretedto meanthat young people are
staying single longer. Because of cohabitation,
however, being unmarriedis not synonymous
with being single. Young people are setting up
housekeepingwith partnersof the oppositesex at
almost as early an age as they did before marriage
rates declined. Three-quarters of the decline in the
proportion of women married for the first time by
TABLE 11. ATTITUDES TOWARD MARRIAGE AMONG NEVER-MARRIED PERSONS UNDER AGE 35: UNADJUSTED PROPO
PROPORTIONS ADJUSTED FOR OTHER VARIABLES BY MULTIPLE CLASSIFICATION ANALYSIS
Parents Prefer
Marrieda
Unadjusted Adjusted
Variable
Freedom Worse if
Marriedb
Unadjusted Adjusted
Marriage Better
than Singlec
Unadjusted Adjusted
Better Childbear
than Childless
Unadjusted Adjus
Cohabiting
Previously married
Never married
36%
40
24%
37
21%
24
24%
25
29%
36
30%
32
30%
34
29%
33
Not cohabiting
Previously married
Never married
42
27
29
29
37
45
42
44
32
34
34
34
30
28
30
28
Education
< 12 years
12 years
College 1-3
College 1-3
College 4 +
College 4+
36
31
34
13
40
35
34
31
32
20
34
31
34
39
44
45
47
45
37
39
45
43
46
35
26
29
35
33
36
52
40
30
35
33
36
52
41
30
28
24
18
38
38
30
28
25
20
40
Sex
Males
Females
29
31
29
31
49
31
49
31
37
30
37
30
30
28
30
27
Race /ethnicity
Blacks
Whites
Mexican Americans
30
28
46
28
28
43
36
43
37
38
42
37
31
32
48
31
32
49
35
24
48
34
25
46
Age
Under 25
25-29
30-34
19
44
49
20
42
47
42
42
37
41
42
43
33
37
33
33
36
33
28
30
29
29
30
29
Total
years
years, enrolled
years
years, enrolled
30%
41%
34%
29%
aProportion responding "Strongly agree" or "Agree" on a 5-point scale to "My parents would like it better if I were marr
bProportion responding "Much worse" or "Somewhat worse" on a 5-point scale to "For each of the following areas, please c
might be different if you were MARRIED now ... (d) freedom to do what you want."
cProportion responding "Strongly agree" or "Agree" on a 5-point scale to "It's better for a person to get married than to g
dProportion responding "Strongly agree" or "Agree" on a 5-point scale to "It's better for a person to have a child than to
eProportion responding "Strongly agree" or "Agree" on a 5-point scale to "Please indicate how much you agree or disagre
statements: ... (d) It would be all right for me to have children without being married-If I had definite plans to marry the fa
926
REFERENCES
927
CohabitationandDecliningRatesof Marriage
Sweet, James A., Larry L. Bumpass, and Vaughn R. A.
Call. 1988. "The design and content of the National
Survey of Families and Households." NSFH Working Paper No. 1, Center for Demography and