Number 323 + May 31, 2001

First Marriage Dissolution, Divorce, and Remarriage:

United States
Matthew D. Bramlett, Ph.D., and William D. Mosher, Ph.D., Division of Vital Statistics

Abstract Introduction
Objectives—This report presents national estimates of the duration of first and Marriage is associated with a
second marriages for women 15–44 years of age in 1995. National estimates of the variety of positive outcomes, and
probability of divorce given separation and of the probability of remarriage given dissolution of marriage is associated
divorce are also presented. with negative outcomes for men,
Methods—The life- table estimates are based on a nationally representative women, and their children. A full
sample of women 15–44 years of age in the United States in 1995 from the analysis of the benefits of marriage to
National Survey of Family Growth, Cycle 5. either children or spouses is beyond the
Results—One fifth of first marriages end within 5 years and one third end scope of this report, but this brief
within 10 years. First marriages of teenagers disrupt faster than the first marriages introduction should serve to highlight
of women who were ages 20 years and older at marriage. First marriages of black the importance of the data described in
non-Hispanic women dissolve at a faster rate: 47 percent end within 10 years this report. This report releases estimates
compared with 34 percent for Hispanic, 32 percent for white non-Hispanic, and of the patterns of marriage, divorce, and
20 percent for Asian non-Hispanic women. Virtually all separations among white remarriage in the United States as of
non-Hispanic women (98 percent) end in divorce within 6 years, compared with 1995 by several important demographic
only 80 percent of separations among Hispanic women and 72 percent of characteristics. A later report will show
separations among black non-Hispanic women. more detailed estimates by a wide
Women under age 25 years at divorce are more likely to remarry than women variety of other characteristics.
at least age 25 years at divorce. White non-Hispanic and Hispanic women are much Compared with unmarried people,
more likely to remarry than black non-Hispanic women. White non-Hispanic married men and women tend to have
women are slightly more likely than Hispanic women to remarry. The data suggest lower mortality, less risky behavior,
that women who remarry before age 25 years are more likely to experience a more monitoring of health, more
second marital disruption than women who remarry at ages older than 25 years, compliance with medical regimens,
although the difference is only significant at late marital durations. Black non- higher sexual frequency, more
Hispanic remarriages are more likely to disrupt than Hispanic or white non-Hispanic
remarriages.
Keywords: marriage c separation c divorce c remarriage

Acknowledgments
The 1995 National Survey of Family Growth was jointly planned and funded primarily by the National Center for Health Statistics (NCHS), the
National Institute for Child Health and Human Development (NICHD), the Office of Population Affairs (OPA), and the National Center for HIV,
STD and TB Prevention (NCHSTP), with additional support from the Children’s Bureau.
The authors gratefully acknowledge the technical assistance of Wayne E. Johnson, Ph.D., of the NCHS Office of Research and Methodology
(ORM) for assistance in estimating standard errors of the statistics in this report.
The authors also gratefully acknowledge the helpful review and comments of Christine Bachrach, Ph.D. (NICHD). This report was edited by
Klaudia Cox and typeset by Annette F. Holman of the Publications Branch, Division of Data Services.
Number 323 + May 31, 2001

First Marriage Dissolution, Divorce, and Remarriage:

United States
Matthew D. Bramlett, Ph.D., and William D. Mosher, Ph.D., Division of Vital Statistics

Abstract Introduction
Objectives—This report presents national estimates of the duration of first and Marriage is associated with a
second marriages for women 15–44 years of age in 1995. National estimates of the variety of positive outcomes, and
probability of divorce given separation and of the probability of remarriage given dissolution of marriage is associated
divorce are also presented. with negative outcomes for men,
Methods—The life- table estimates are based on a nationally representative women, and their children. A full
sample of women 15–44 years of age in the United States in 1995 from the analysis of the benefits of marriage to
National Survey of Family Growth, Cycle 5. either children or spouses is beyond the
Results—One fifth of first marriages end within 5 years and one third end scope of this report, but this brief
within 10 years. First marriages of teenagers disrupt faster than the first marriages introduction should serve to highlight
of women who were ages 20 years and older at marriage. First marriages of black the importance of the data described in
non-Hispanic women dissolve at a faster rate: 47 percent end within 10 years this report. This report releases estimates
compared with 34 percent for Hispanic, 32 percent for white non-Hispanic, and of the patterns of marriage, divorce, and
20 percent for Asian non-Hispanic women. Virtually all separations among white remarriage in the United States as of
non-Hispanic women (98 percent) end in divorce within 6 years, compared with 1995 by several important demographic
only 80 percent of separations among Hispanic women and 72 percent of characteristics. A later report will show
separations among black non-Hispanic women. more detailed estimates by a wide
Women under age 25 years at divorce are more likely to remarry than women variety of other characteristics.
at least age 25 years at divorce. White non-Hispanic and Hispanic women are much Compared with unmarried people,
more likely to remarry than black non-Hispanic women. White non-Hispanic married men and women tend to have
women are slightly more likely than Hispanic women to remarry. The data suggest lower mortality, less risky behavior,
that women who remarry before age 25 years are more likely to experience a more monitoring of health, more
second marital disruption than women who remarry at ages older than 25 years, compliance with medical regimens,
although the difference is only significant at late marital durations. Black non- higher sexual frequency, more
Hispanic remarriages are more likely to disrupt than Hispanic or white non-Hispanic
remarriages.
Keywords: marriage c separation c divorce c remarriage

Acknowledgments
The 1995 National Survey of Family Growth was jointly planned and funded primarily by the National Center for Health Statistics (NCHS), the
National Institute for Child Health and Human Development (NICHD), the Office of Population Affairs (OPA), and the National Center for HIV,
STD and TB Prevention (NCHSTP), with additional support from the Children’s Bureau.
The authors gratefully acknowledge the technical assistance of Wayne E. Johnson, Ph.D., of the NCHS Office of Research and Methodology
(ORM) for assistance in estimating standard errors of the statistics in this report.
The authors also gratefully acknowledge the helpful review and comments of Christine Bachrach, Ph.D. (NICHD). This report was edited by
Klaudia Cox and typeset by Annette F. Holman of the Publications Branch, Division of Data Services.
2 Advance Data No. 323 + May 31, 2001
satisfaction with their sexual lives, more levels of parental involvement in school similar to that of white Americans as
financial savings, and higher wages activities and lower student achievement measured by attitudes towards marriage
(1–3). The differences between married (10). Children raised in single-parent (14,15). One prominent explanation
and unmarried people may reflect a families are more likely to drop out of offered by some researchers for the
causal effect of marriage or a selection high school, have lower grades and lower proportion of time spent in
effect (healthier people may be more attendance while in school, and are less marriage among black Americans is the
likely than others to find mates and to likely to attend and graduate from idea of a ‘‘marriage squeeze,’’ in which
marry). Research has suggested that the college than children raised in two- the ‘‘marriageable pool’’ of black men is
benefits of marriage may be partially parent families (11). They are more low due to high rates of joblessness,
due to a selection effect and partially likely to be out of school and incarceration, and mortality (16–18).
due to true benefits to be gained from unemployed and are more likely to Employed men are more likely than
being married as opposed to being become single parents than children unemployed men to marry (19).
unmarried (3,4). A lower mortality risk raised in two-parent families (11). In addition to race and employment
among those who are married has been Studies have found that compared with status, other characteristics of
shown to persist even after health in children in two-parent families, children individuals that have been found to be
early adulthood was controlled. This of divorce score lower on measures of related to the probability of marriage
suggests that at least part of the benefit self-concept, social competence, include education and earnings (20),
of being married is not the result of conduct, psychological adjustment, and intact status of family of origin, and
selection (4). long-term health (5). parents’ educational levels (21). Other
Compared with married individuals, The positive health benefits of characteristics of individuals related to
divorcees exhibit lower levels of marriage and the negative consequences the probability of divorce include age at
psychological well-being, more health of divorce illustrate the importance of marriage, education, birth cohort (22),
problems and greater risk of mortality, examining trends and differentials in the religion, marriage cohort, fertility status
more social isolation, less satisfying sex patterns of marriage and divorce over at marriage (23), premarital cohabitation
lives, more negative life events, greater time. (24), and premarital sexual activity (25).
levels of depression and alcohol use, Other characteristics related to the
and lower levels of happiness and Trends and differences in probability of remarriage include
self-acceptance (5). The economic marriage and divorce education and age at divorce (26) and
consequences of divorce can be severe presence of children from prior
for women. Most often, children remain In the United States in the second marriages (9,22).
with the mother after divorce; the loss half of the twentieth century, the The lower economic prospects of
of the ex-husband’s income often results proportion of people’s lives spent in less-educated young men has been
in a severe loss of income per capita marriage declined due to postponement hypothesized to decrease the probability
(6,7). For men, the retention of income of marriage to later ages, greater of marriage. The increasing economic
combined with decreased family size incidence of never marrying, and higher independence of women has also been
may actually result in an increase in his rates of divorce. The increase in hypothesized to decrease the probability
new household’s income per capita nonmarital cohabiting has also of marriage, although recent evidence
(6,8). contributed to the decline in the suggests that the increasing economic
Adverse outcomes accrue to proportion of people’s lives spent in independence of women may actually
children of divorce and children raised marriage. Increasing rates of increase the probability of marriage
in single-parent families. Although not cohabitation have largely offset because earnings and employment may
all single-parent families are the result decreasing rates of marriage (12,13). make either partner an attractive
of divorce and not all divorced mothers The proportion of time spent in potential spouse (16,20). Marriage
remain single, virtually all children of marriage has varied across demographic market conditions may also play a role
divorce spend some time in a single- subgroups. Since 1950, the marital in that the probability of divorce is
parent household until the mother patterns of white and black Americans higher in areas with large numbers of
remarries. Even when the mother does have diverged considerably. About economically attractive potential
remarry, studies suggest that children in 91 percent of white women born in the alternate partners (16).
stepfamilies are similar to children in 1950’s are estimated to marry at some A full analysis of all the individual
single-parent families: both groups of time in their lives, compared with only and community-level characteristics
children do worse than children living 75 percent of black women born in the associated with marriage and divorce is
with two parents in terms of academic 1950’s (12). Black married couples are beyond the scope of this report. This
achievement, depression, and behavior more likely to break up than white report releases estimates of the patterns
problems such as drug and alcohol married couples, and black divorcees are of marriage, divorce, and remarriage in
abuse, premarital sexual intercourse, and less likely to remarry than white the United States as of 1995 by a few
being arrested (9). divorcees (12). important demographic characteristics.
Compared with two-parent families, The degree of attachment to A later report will show more detailed
single-parent families demonstrate lower marriage among black Americans is
 Advance Data No. 323 + May 31, 2001 3
estimates by a wide variety of other including the beginning and ending As members of the cohort age, they
characteristics. dates of each marriage and the outcome are subjected to the age-specific death
of each marriage (separation, divorce, rates of successive age categories in the
Methods widowhood) (29). Given these data, the life table. At each interval, the age-
probabilities of first marriage disruption, specific death rate for that interval is
Data of divorce after separation, of used to calculate how many members of
The national estimates of marriage remarriage after divorce, and of second the cohort die during that interval. That
and divorce patterns in this report are marriage disruption can be estimated number of deaths is subtracted from the
based on data from the 1995 National using life table techniques. count of cohort members, and the result
Survey of Family Growth (NSFG). Previous analyses of marriage and is the number of cohort members who
Cycle 5 of the NSFG, conducted by the divorce based on vital statistics have survive and are counted in the next
Centers for Disease Control and computed and presented rates of interval. Eventually, the last age interval
Prevention’s National Center for Health marriage and divorce (30, 31). Rates are is reached and the last cohort members
Statistics (NCHS), was based on a snapshots of data limited to specific die. One overall measure of longevity is
multi-stage probability sample of the time points. The life table analysis in the proportion who survive to specific
civilian noninstitutionalized population this report takes a life-cycle approach to ages (33). Survivor curves can be
of women in the United States. This estimate the probabilities that: plotted that show the proportion of the
cycle yielded estimates that are (a) a first marriage will end in cohort surviving to each successive age
representative of women 15–44 years of separation or divorce, category (34,35).
age in 1995. Between January and (b) a separation will result in To apply life table analysis to the
October 1995, in-home computer- divorce, study of marital stability, the cohort of
assisted personal interviews (CAPI) (c) a divorce from a first marriage people is replaced with a cohort of
were conducted with 10,847 women, of will be followed by remarriage, and marriages; age is replaced by marital
whom 1,553 were Hispanic women, (d) a second marriage will end in duration, and death is replaced by
6,483 were white non-Hispanic women, separation or divorce. separation or divorce. In addition, one
2,446 were black non-Hispanic women, other issue must be addressed. The
Previous analysis of divorce and
and 365 were women of other races and sample of women is limited to ages
remarriage based on cycle 4 of the
ethnic origins. The overall response rate 15–44 years, so the marriage histories
National Survey of Family Growth used
was 79 percent (27). are incomplete. For respondents whose
a measure of the cumulative proportion
The sample list for the 1995 NSFG marriage has not yet ended as of the
of marriages disrupted as of interview to
was selected from households that time of the interview, the end date of
describe the phenomena (32). An
responded to the 1993 National Health the marriage is unknown. Therefore, the
improvement over a rate, this statistic
Interview Survey (NHIS). Black and duration of the marriage is unknown and
approximates the estimates that life table
Hispanic women were sampled at higher is referred to in statistical literature as
analysis provides. However, it is only a
rates than were other women. Sampling ‘‘censored.’’ Life table procedures allow
single measure of the cumulative
weights account for differential for the simultaneous analysis of both
proportion at the time of interview; life
probabilities of sample selection and for complete and incomplete marriage
tables provide estimates of cumulative
nonresponse and are adjusted to agree histories (22).
proportions at every time point in the
with control totals by age, race, parity, Life table analysis can handle
life course of a marriage.
and marital status provided by the U.S. censored cases by keeping such cases in
Census Bureau. The 10,847 women in Life tables the analysis as long as they are at risk
the 1995 NSFG represent the 60 million of disruption and then dropping them
The life table is a tool that out once the risk is unknown (36). For
women 15–44 years of age in the
demographers most often use to study example, when calculating the
civilian noninstitutionalized population
mortality, but it is often applied to the proportion of marriages that dissolve in
of the United States in 1995. On
study of marital stability. In studying each duration interval, a marriage that
average, each woman in the 1995 NSFG
mortality, the cohort life table is a has existed for 24 months and is still
represents about 5,500 women in the
summary of the mortality history of a intact at the time of the interview would
population, although sample weights
given cohort from birth to death and remain in the denominator for each
vary considerably from this average
requires data on the longevity of all duration interval until 24 months of
value depending on the respondent’s
cohort members, a span of more than duration is reached, at which point the
race, age, and Hispanic ethnicity; the
100 years. As a result, the period life case would no longer be used in the
response rate for similar women; and
table is typically used as a model of calculations.
other factors (27,28). See the Technical
what would happen to a given cohort if Widowhood removes a marriage
notes for additional information.
the age-specific death rates from a from the risk of dissolution. The length
The 1995 NSFG collected complete
certain point in time were to remain of time that the marriage would have
retrospective histories of each woman’s
fixed for the duration of the cohort’s life endured intact if the husband had not
experiences with marriage and divorce,
(33,34). died is unknown, so cases of
4 Advance Data No. 323 + May 31, 2001
widowhood are censored (removed from Table 1. Number and percent distribution of women 15–44 years of age by age at
interview and race/ethnicity, according to past cohabitation and marital status: United
the analysis) at the date of the States, 1995
husband’s death. Widowhood is very
rare in the age group 15–44 years. The Past cohabitation and marital status

mortality of wives is unobservable Never married Ever married

because the woman had to have been
Age at interview Number Never Ever Never Ever
alive to be interviewed. As the risk of and race/ethnicity in thousands Total cohabited cohabited cohabited cohabited
mortality among women in the age
range 15–44 years is low, this is Total. . . . . . . . . . . . . . . . . . . . . 60,201 100.0 27.5 10.2 31.4 30.9

unlikely to substantially affect the Age at interview

results. 15–19 years . . . . . . . . . . . . . . . . 8,961 100.0 88.6 7.0 2.6 1.9
The probability of divorce is not 20–24 years . . . . . . . . . . . . . . . . 9,041 100.0 45.5 20.2 16.2 18.1
25–29 years . . . . . . . . . . . . . . . . 9,693 100.0 20.3 15.4 30.4 33.9
always the best measure of marital
30–34 years . . . . . . . . . . . . . . . . 11,065 100.0 10.8 9.3 37.8 42.1
instability. Subgroup comparisons of the 35–39 years . . . . . . . . . . . . . . . . 11,211 100.0 7.1 6.4 42.9 43.6
probability of divorce are not 40–44 years . . . . . . . . . . . . . . . . 10,230 100.0 5.5 4.1 51.5 38.9
appropriate for subgroups that differ in Race/ethnicity
the conditional probability of divorce Hispanic . . . . . . . . . . . . . . . . . . 6,702 100.0 28.2 10.4 35.1 26.3
given separation (37). Research has White non-Hispanic . . . . . . . . . . . 42,522 100.0 24.7 8.9 32.7 33.7
shown that the marriages of black Black non-Hispanic . . . . . . . . . . . 8,210 100.0 39.7 17.3 20.2 22.9
women are more likely to end in Other non-Hispanic . . . . . . . . . . . 2,767 100.0 33.2 8.3 35.1 23.4

separation than the marriages of white

women, and that separated black couples
are less likely to make the transition to
report, Asian non-Hispanic women are marriage, and black non-Hispanic
divorce than separated white couples
included in the analysis of the full women are more likely to have
(22,32). Therefore, a comparison of the
sample, but omitted from separate experienced neither cohabitation nor
probability of divorce alone obscures
analysis (see Technical notes). marriage (table 1).
some of the difference between these
The statistics in this report were Table 2 shows the distribution of
two groups in the probability that a
computed using the Lifetest procedure women 15–44 years of age in 1995, by
marriage will disrupt. For this reason, in
in version 8 of PC-SAS (38). current cohabitation and marital status at
this report, marital disruption is defined interview, age at interview, and race/
as either separation or divorce, and a
Results ethnicity. Current cohabitation and
second analysis examines the probability marital status is categorized as currently
of divorce, given separation. Cohabitation and marital status cohabiting or not currently cohabiting.
The Technical notes include an The category not currently cohabiting is
example of a life table for the duration Table 1 shows the distribution of
further split into never married, formerly
of first marriage and a description of women 15–44 years of age in 1995 by
married, or currently married. These
each part of the life table and its role in past cohabitation and marital status, age
four subgroups are mutually exclusive
the generation of survival statistics. In at interview, and race/ethnicity. Past
and exhaustive, summing to 100 percent.
the analysis below, for the sake of cohabitation and marital status is
Roughly 50 percent of women ages
brevity, only the cumulative proportion categorized as never married or ever
15–44 years are currently married, and
dissolved at the beginning of each married, with each group further split
7 percent of women ages 15–44 years
interval is presented and compared into two subgroups separating the never
are currently cohabiting (table 2). One
across subgroups. In addition, the cohabited from the ever cohabited.
third of women ages 15–44 years are
overall percent of cases censored is These four subgroups are mutually
not cohabiting and have never married.
presented for each subgroup as a exclusive and exhaustive, summing to
The remaining 10 percent are not
comparative measure of the overall 100 percent.
cohabiting and are formerly married.
likelihood of disruption. In analyses of Almost 28 percent of women ages
The percent currently cohabiting is
the probability of marital disruption, 15–44 years have no union experience,
larger for young adults in their twenties,
subgroups with a relatively high percent having never married nor cohabited
and then decreases as age increases. The
censored by interview are less likely to (table 1). This percent is considerably
percent never married decreases as age
disrupt (are more likely to have larger for young women and decreases
increases, while the percent formerly
remained intact until interview). as age increases. About 62 percent of
married and the percent currently
The analysis of first marriage women have ever been married, half of
married increase as age increases. The
stability is the only analysis in this whom have ever cohabited. The
most striking differences by race/
report in which there were sufficient remaining 10 percent have cohabited,
ethnicity are the higher percent not
numbers of Asian non-Hispanic women but never married. White non-Hispanic
cohabiting and never married and the
in the NSFG sample to generate reliable women are more likely to have
estimates. In all other analyses in this experienced both cohabitation and
 Advance Data No. 323 + May 31, 2001 5
Table 2. Number and percent distribution of women 15–44 years of age by age at lower percent currently married among
interview and race/ethnicity, according to current cohabitation and marital status:
United States, 1995
black non-Hispanic women.
The focus of this report is on
Current cohabitation and marital status marriage. A more detailed analysis of
Not cohabiting cohabitation is beyond the scope of this
report. A later report will include
Age at interview Number Currently Never Formerly Currently
and race/ethnicity in thousands Total cohabiting married married married analysis of cohabitation, including
analysis of the probability of
Total. . . . . . . . . . . . . . . . . . . 60,201 100.0 7.0 33.4 10.3 49.3 cohabitational dissolution, the
Age at interview probability of a cohabitation making the
15–19 years . . . . . . . . . . . . . . 8,961 100.0 4.1 91.5 0.6 3.8 transition to marriage, and the
20–24 years . . . . . . . . . . . . . . 9,041 100.0 11.2 56.1 5.5 27.2 probability of cohabitation after the
25–29 years . . . . . . . . . . . . . . 9,693 100.0 9.8 28.9 8.8 52.5
dissolution of the first marriage.
30–34 years . . . . . . . . . . . . . . 11,065 100.0 7.5 16.2 11.6 64.7
35–39 years . . . . . . . . . . . . . . 11,211 100.0 5.3 11.9 15.0 67.9
40–44 years . . . . . . . . . . . . . . 10,230 100.0 4.4 8.8 18.1 68.6 First marriage
Race/ethnicity After 5 years, 20 percent of all first
Hispanic . . . . . . . . . . . . . . . . 6,702 100.0 8.2 32.8 11.6 47.4 marriages have disrupted, due to
White non-Hispanic . . . . . . . . . 42,522 100.0 7.0 29.4 9.3 54.3 separation or divorce (table 3). Data
Black non-Hispanic . . . . . . . . . 8,210 100.0 6.9 52.5 15.5 25.2
from the former divorce registration
Other non-Hispanic . . . . . . . . . 2,767 100.0 4.6 39.1 7.6 48.8
system reported only divorces, not
separations (31). While 26.5 percent of
women have divorced at the end of 10
years of marriage, 33 percent of all first
marriages have disrupted due to either
Table 3. Probability of first marriage disruption by duration of marriage and wife’s age at separation or divorce at the end of 10
marriage: United States, 1995 years (NSFG cycle 5 results not shown).
Age Separation and divorce are both
classified as marital dissolution because
Proportion of first marriages Less than 18–19 20–24 25 years
disrupted1 after— Total 18 years years years and over different groups have differential
probabilities of making the transition
0 months . . . . . . . . . . . . . . . . . . . . . . . – – – – –
from separation to divorce (22,37).
6 months . . . . . . . . . . . . . . . . . . . . . . . 0.01 0.02 0.01 0.01 0.01
12 months . . . . . . . . . . . . . . . . . . . . . . 0.03 0.04 0.04 0.02 0.02 First marriages by wife’s age at
18 months . . . . . . . . . . . . . . . . . . . . . . 0.05 0.07 0.06 0.04 0.04 marriage—If the wife was a teenager at
24 months . . . . . . . . . . . . . . . . . . . . . . 0.07 0.10 0.09 0.06 0.05 first marriage, the marriage is much
30 months . . . . . . . . . . . . . . . . . . . . . . 0.09 0.13 0.12 0.08 0.06
more likely to dissolve than if the wife
36 months . . . . . . . . . . . . . . . . . . . . . . 0.12 0.17 0.15 0.10 0.08
42 months . . . . . . . . . . . . . . . . . . . . . . 0.14 0.20 0.18 0.12 0.10 was at least 20 years of age at marriage.
48 months . . . . . . . . . . . . . . . . . . . . . . 0.16 0.23 0.20 0.14 0.11 The first marriages of brides under age
54 months . . . . . . . . . . . . . . . . . . . . . . 0.18 0.27 0.22 0.16 0.13 18 years are the most likely to dissolve
60 months . . . . . . . . . . . . . . . . . . . . . . 0.20 0.29 0.24 0.17 0.15
72 months . . . . . . . . . . . . . . . . . . . . . . 0.23 0.34 0.28 0.19 0.17
at all marital durations, followed by the
84 months . . . . . . . . . . . . . . . . . . . . . . 0.26 0.39 0.32 0.22 0.19 marriages of women 18–19 years of age
96 months . . . . . . . . . . . . . . . . . . . . . . 0.29 0.42 0.35 0.25 0.20 at marriage. Marriages of women at
108 months . . . . . . . . . . . . . . . . . . . . . 0.31 0.45 0.37 0.27 0.22 least age 20 years at marriage are much
120 months . . . . . . . . . . . . . . . . . . . . . 0.33 0.48 0.40 0.29 0.24
132 months . . . . . . . . . . . . . . . . . . . . . 0.35 0.49 0.42 0.31 0.26
less likely to dissolve (table 3 and
144 months . . . . . . . . . . . . . . . . . . . . . 0.38 0.52 0.44 0.33 0.29 figure 1). After 10 years of marriage,
156 months . . . . . . . . . . . . . . . . . . . . . 0.39 0.54 0.46 0.33 0.30 48 percent of marriages of women under
168 months . . . . . . . . . . . . . . . . . . . . . 0.41 0.57 0.47 0.35 0.31 age 18 years at marriage have disrupted
180 months . . . . . . . . . . . . . . . . . . . . . 0.43 0.59 0.49 0.36 0.35
192 months . . . . . . . . . . . . . . . . . . . . . 0.44 0.60 0.51 0.37 2
0.36
compared with 40 percent of marriages
204 months . . . . . . . . . . . . . . . . . . . . . 0.46 0.62 0.53 0.38 2
0.43 of women who were 18–19 years of age
2
216 months . . . . . . . . . . . . . . . . . . . . . 0.47 0.63 0.54 0.39 0.43 at marriage, 29 percent of marriages of
2
228 months . . . . . . . . . . . . . . . . . . . . . 0.48 0.65 0.55 0.40 0.43
2
women who were 20–24 years of age at
240 months . . . . . . . . . . . . . . . . . . . . . 0.50 0.67 0.56 0.41 –––
marriage, and 24 percent of marriages of
Percent censored (intact at interview) . . . . . 66.00 41.90 53.10 71.70 81.50
women at least 25 years of age at
– Quantity zero. marriage. The difference in the
– – – Data not available.
1
Disruption refers to either separation or divorce.
probability of disruption between
2
Estimates may be affected by bias resulting from incomplete marriage histories and should be interpreted with caution. The women ages 20–24 years and women at
durations in subsequent tables are limited to avoid this possible bias. For details, see section on ‘‘First marriages by wife’s age
at marriage.’’ least 25 years of age at marriage is not
NOTE: This table includes all first marriages of women 15–44 years of age in 1995. statistically significant, but the difference
6 Advance Data No. 323 + May 31, 2001
between those over age 20 years and
Under 18 years 18 -19 years
those under age 20 years at marriage is
20 - 24 years 25 years and over
70 significant.
As noted earlier, a larger percent
censored means that more marriages
60
were still intact at the time of the
interview. Due to the age limitation on
Percent of first marriages disrupted

the sample (15–44 years old), the older

50
the age at marriage, the shorter the
average time span available between
40 marriage and interview, and the less
likely that enough time has passed for
marital dissolution to occur. Thus, the
30 percent censored increases with age at
marriage (table 3).
A woman 30 years of age at the
20 time of her marriage cannot be included
in a measure of the probability of
dissolution after 20 years of marriage
10
because she would have been 50 years
of age at 20 years’ duration of marriage,
0 and the maximum age of women in the
0 2 4 6 8 10 12 14 16 18 20 NSFG sample was 44 years old. Due to
Duration in years the age limitation on the sample, the
longer the period of observation, the
Figure 1. Probability of first marriage disruption by duration of marriage and wife’s age at
younger the women must be at marriage
marriage: United States, 1995
to fit within the duration limits.
Estimates about the later durations are

Table 4. Probability of first marriage disruption by duration of marriage and wife’s race/ethnicity: United States, 1995

Proportion of first marriages White Black Asian

disrupted1after— Total non-Hispanic non-Hispanic non-Hispanic Hispanic

0 months . . . . . . . . . . . . . . . . . . . . . . . – – – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . 0.01 0.01 0.02 0.01 0.01
12 months . . . . . . . . . . . . . . . . . . . . . . 0.03 0.03 0.05 0.01 0.03
18 months . . . . . . . . . . . . . . . . . . . . . . 0.05 0.05 0.08 0.02 0.05
24 months . . . . . . . . . . . . . . . . . . . . . . 0.07 0.07 0.11 0.03 0.06
30 months . . . . . . . . . . . . . . . . . . . . . . 0.09 0.09 0.13 0.04 0.09
36 months . . . . . . . . . . . . . . . . . . . . . . 0.12 0.12 0.15 0.06 0.11
42 months . . . . . . . . . . . . . . . . . . . . . . 0.14 0.14 0.19 0.07 0.13
48 months . . . . . . . . . . . . . . . . . . . . . . 0.16 0.16 0.23 0.07 0.14
54 months . . . . . . . . . . . . . . . . . . . . . . 0.18 0.18 0.25 0.08 0.16
60 months . . . . . . . . . . . . . . . . . . . . . . 0.20 0.20 0.28 0.10 0.17
72 months . . . . . . . . . . . . . . . . . . . . . . 0.23 0.22 0.33 0.12 0.22
84 months . . . . . . . . . . . . . . . . . . . . . . 0.26 0.25 0.38 0.15 0.26
96 months . . . . . . . . . . . . . . . . . . . . . . 0.29 0.28 0.42 0.16 0.28
108 months . . . . . . . . . . . . . . . . . . . . . 0.31 0.30 0.44 0.18 0.32
120 months . . . . . . . . . . . . . . . . . . . . . 0.33 0.32 0.47 0.20 0.34
132 months . . . . . . . . . . . . . . . . . . . . . 0.35 0.34 0.49 0.23 0.35
144 months . . . . . . . . . . . . . . . . . . . . . 0.38 0.36 0.51 0.23 0.37
156 months . . . . . . . . . . . . . . . . . . . . . 0.39 0.38 0.53 0.23 0.38
168 months . . . . . . . . . . . . . . . . . . . . . 0.41 0.39 0.54 0.23 0.40
180 months . . . . . . . . . . . . . . . . . . . . . 0.43 0.42 0.55 0.23 0.42
192 months . . . . . . . . . . . . . . . . . . . . . 0.44 0.43 0.57 0.23 0.42
204 months . . . . . . . . . . . . . . . . . . . . . 0.46 0.45 0.60 0.27 0.46
216 months . . . . . . . . . . . . . . . . . . . . . 0.47 0.46 0.61 0.27 0.48
228 months . . . . . . . . . . . . . . . . . . . . . 0.48 0.47 0.61 0.27 0.49
2 2 2 2 2
240 months . . . . . . . . . . . . . . . . . . . . . 0.50 0.48 0.63 0.31 0.52
Percent censored (intact at interview) . . . . . 66.00 66.30 56.00 83.10 67.90

– Quantity zero.
1
Disruption refers to either separation or divorce.
2
Estimates may be affected by bias resulting from incomplete marriage histories and should be interpreted with caution. The durations in subsequent tables are limited to avoid this possible bias.
For details, see section on ‘‘First marriages by wife’s age at marriage.’’
NOTE: This table includes all first marriages of women 15–44 years of age in 1995.
 Advance Data No. 323 + May 31, 2001 7
affected by this limitation, tables and
70 Black non-Hispanic White non-Hispanic graphs in this report are truncated as
Asian non-Hispanic Hispanic
necessary. The events examined in this
report include first marriage dissolution,
60
the transition from separation to divorce,
second marriage, and second marriage
dissolution. The higher the average age
Percent of first marriages disrupted

50 at the event, the more truncation is

necessary to avoid this potential bias. In
the future, the NSFG could address this
40
issue by interviewing women up to 54
or 59 years of age.
First marriages by wife’s race/
30
ethnicity—The probability of first
marital disruption varies by race/
ethnicity. After 10 years, 32 percent of
20
first marriages of white non-Hispanic
women have dissolved, and 34 percent
10
of first marriages of Hispanic people
have dissolved (table 4). In sharp
contrast, 47 percent of first marriages
0
among black non-Hispanic women have
0 2 4 6 8 10 12 14 16 18 20 dissolved after 10 years. First marriages
Duration in years of Asian non-Hispanic women dissolve
at a considerably slower rate: after 10
Figure 2. Probability of first marriage disruption by duration of marriage and wife’s years, only 20 percent have disrupted
race/ethnicity: United States, 1995
(table 4 and figure 2).
The lower percent censored for
therefore biased towards the experiences this means that estimates towards the
black non-Hispanic women in table 4 is
of younger women at marriage. Because later durations may be overestimates of
a reflection of their lower likelihood of
younger age at marriage is associated the probability of disruption. To avoid
remaining married until the time of the
with a higher probability of disruption, awkwardness in describing results

Table 5. Probability of first marriage separations making the transition to divorce by duration of separation and wife’s age at separation:
United States, 1995

Age
Proportion of first separations
that have made the transition Less than 20–24 25 years
to divorce after— Total 20 years years and over

0 months . . . . . . . . . . . . . . . . . . . . . . . – – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . 0.32 0.41 0.33 0.29
12 months . . . . . . . . . . . . . . . . . . . . . . 0.54 0.58 0.57 0.51
18 months . . . . . . . . . . . . . . . . . . . . . . 0.69 0.70 0.72 0.66
24 months . . . . . . . . . . . . . . . . . . . . . . 0.75 0.74 0.77 0.73
30 months . . . . . . . . . . . . . . . . . . . . . . 0.80 0.82 0.82 0.79
36 months . . . . . . . . . . . . . . . . . . . . . . 0.84 0.84 0.86 0.83
42 months . . . . . . . . . . . . . . . . . . . . . . 0.86 0.86 0.88 0.85
48 months . . . . . . . . . . . . . . . . . . . . . . 0.88 0.87 0.89 0.87
54 months . . . . . . . . . . . . . . . . . . . . . . 0.89 0.89 0.91 0.89
60 months . . . . . . . . . . . . . . . . . . . . . . 0.91 0.89 0.92 0.90
72 months . . . . . . . . . . . . . . . . . . . . . . 0.92 0.92 0.93 0.92
84 months . . . . . . . . . . . . . . . . . . . . . . 0.93 0.93 0.94 0.92
96 months . . . . . . . . . . . . . . . . . . . . . . 0.94 0.93 0.95 0.93
108 months . . . . . . . . . . . . . . . . . . . . . 0.94 0.93 0.95 0.94
120 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.93 0.96 0.95
132 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.94 0.96 0.96
144 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.94 0.96 0.96
156 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.94 0.96 0.96
168 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.94 0.96 0.96
180 months . . . . . . . . . . . . . . . . . . . . . 0.96 0.94 0.96 0.97
Percent censored (not yet divorced as of
interview) . . . . . . . . . . . . . . . . . . . . . . 13.60 9.10 8.70 18.00

– Quantity zero.
NOTE: This table includes all separations from first marriages of women 15–44 years of age in 1995.
8 Advance Data No. 323 + May 31, 2001
interview, and the higher percent
Less than 20 years 20- 24 years 25 years and over censored for Asian non-Hispanic women
100
reflects their higher likelihood of
remaining married until interview.
Percent of separated women who have divorced

Divorce following separation

80
There are no statistically significant
differences by age at separation in the
probability of divorce given separation
60 (table 5 and figure 3). Figure 3 shows
that the probability of divorce is
virtually identical across all age groups
at all durations of separation. Most
separated women make the transition to
40
divorce very quickly: 75 percent make
the transition to divorce within 2 years,
and 90 percent do so within 5 years
(table 5).
20 The separations of white non-
Hispanic women are much more likely
to result in divorce than the separations
of black non-Hispanic and Hispanic
0 women (table 6 and figure 4). The
0 2 4 6 8 10 12 14 separations of black non-Hispanic
Duration in years women are the least likely to result in
divorce, but differences between black
Figure 3. Probability of divorce by duration of separation and wife’s age at separation:
United States, 1995 non-Hispanic and Hispanic women are

Table 6. Probability of first marriage separations making the transition to divorce by duration of separation and wife’s race/ethnicity:
United States, 1995

Proportion of first separations

that have made the transition White Black
to divorce after— Total non-Hispanic non-Hispanic Hispanic

0 months . . . . . . . . . . . . . . . . . . . . . . . – – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . 0.32 0.36 0.20 0.22
12 months . . . . . . . . . . . . . . . . . . . . . . 0.54 0.60 0.30 0.40
18 months . . . . . . . . . . . . . . . . . . . . . . 0.69 0.76 0.42 0.52
24 months . . . . . . . . . . . . . . . . . . . . . . 0.75 0.82 0.46 0.56
30 months . . . . . . . . . . . . . . . . . . . . . . 0.80 0.88 0.52 0.62
36 months . . . . . . . . . . . . . . . . . . . . . . 0.84 0.91 0.57 0.66
42 months . . . . . . . . . . . . . . . . . . . . . . 0.86 0.93 0.60 0.69
48 months . . . . . . . . . . . . . . . . . . . . . . 0.88 0.94 0.62 0.71
54 months . . . . . . . . . . . . . . . . . . . . . . 0.89 0.96 0.65 0.73
60 months . . . . . . . . . . . . . . . . . . . . . . 0.91 0.97 0.67 0.77
72 months . . . . . . . . . . . . . . . . . . . . . . 0.92 0.98 0.72 0.80
84 months . . . . . . . . . . . . . . . . . . . . . . 0.93 0.98 0.74 0.81
96 months . . . . . . . . . . . . . . . . . . . . . . 0.94 0.98 0.76 0.82
108 months . . . . . . . . . . . . . . . . . . . . . 0.94 0.98 0.78 0.84
120 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.99 0.79 0.85
132 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.99 0.83 0.85
144 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.99 0.83 0.85
156 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.99 0.83 0.85
168 months . . . . . . . . . . . . . . . . . . . . . 0.95 0.99 0.83 0.85
180 months . . . . . . . . . . . . . . . . . . . . . 0.96 0.99 0.84 0.85
Percent censored (not yet divorced as of
interview) . . . . . . . . . . . . . . . . . . . . . . 13.60 8.90 30.60 25.90

– Quantity zero.
NOTE: This table includes all separations from first marriages of women 15–44 years of age in 1995.
 Advance Data No. 323 + May 31, 2001 9
slight. Virtually all separations of white
White non-Hispanic Hispanic Black non-Hispanic non-Hispanic women result in divorce
100 very quickly: table 6 shows that
91 percent make the transition to divorce
within 3 years of the separation. The
very low percent censored for white
Percent of separated women who have divorced

80 non-Hispanic women shows that there

were very few of them whose
separations had not yet resulted in
divorce by the time of the interview.
Figure 4 shows that for the first 5
60
years of the separation, Hispanic women
are more likely than black non-Hispanic
women to make the transition to divorce
(table 6). After 5 years of separation,
40 differences in the probability of divorce
between separated Hispanic and black
non-Hispanic women are not statistically
significant. A substantial proportion
20 (15 percent or more) of Hispanic and
black non-Hispanic women who are
separated remain separated for the long
term.
0 Remarriage following divorce
0 2 4 6 8 10 12 14
Duration in years Seventy-five percent of divorced
women remarry within 10 years
Figure 4. Probability of divorce by duration of separation and wife’s race/ethnicity:
United States, 1995
(table 7). The probability of remarriage

Table 7. Probability of second marriage by duration of first divorce and woman’s age at divorce: United States, 1995

Age

Proportion of divorced Less than 25 years

women who have remarried after— Total 25 years and over

0 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.09 0.09 0.09
12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.15 0.17 0.14
18 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.23 0.24 0.21
24 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.28 0.28 0.27
30 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.34 0.36 0.33
36 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.39 0.41 0.37
42 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.43 0.45 0.41
48 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.47 0.49 0.44
54 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.51 0.54 0.48
60 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.54 0.57 0.51
72 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.59 0.64 0.55
84 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.62 0.67 0.58
96 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.68 0.72 0.63
108 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.71 0.77 0.65
120 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.75 0.81 0.68
132 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.77 0.83 0.70
144 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.78 0.84 0.72
156 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.80 0.86 0.73
168 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.81 0.87 0.75
180 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.83 0.88 0.76
Percent censored (not yet remarried as of interview) . . . 37.90 24.60 47.60

– Quantity zero.
NOTE: This table includes all first divorces of women 15–44 years of age in 1995.
10 Advance Data No. 323 + May 31, 2001
is significantly higher for women who
100 Less than 25 years 25 years and over were younger at divorce: after 10 years
of divorce, 81 percent of women who
were under age 25 at divorce have
remarried compared with 68 percent of
Percent of divorced women who have remarried

80 women age 25 years and older at

divorce (table 7 and figure 5). The
analysis of remarriage by age at divorce
is constrained to these two age
categories due to the combination of
60
two factors: very small numbers of
women in the sample were under age 20
years at divorce, and the age limitation
on the sample results in small numbers
40 of women over age 30 years at divorce
reaching late durations of divorce.
Black non-Hispanic women are the
least likely to remarry, and white non-
20
Hispanic women are the most likely to
remarry (table 8 and figure 6). The
curves in figure 6 show that the chance
for remarriage for Hispanic women falls
between that of white non-Hispanic and
0 black non-Hispanic women. At early
0 2 4 6 8 10 12 14
durations, the curve for Hispanic women
Duration in years
is closer to that of black non-Hispanic
Figure 5. Probability of remarriage by duration of divorce and woman’s age at divorce: women (and the difference between
United States, 1995 Hispanic and black non-Hispanic
women is not statistically significant).

Table 8. Probability of second marriage by duration of first divorce and woman’s race/ethnicity: United States, 1995

Proportion of divorced White Black

women who have remarried after— Total non-Hispanic non-Hispanic Hispanic

0 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . – – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.09 0.10 0.05 0.05
12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.15 0.17 0.08 0.08
18 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.23 0.25 0.12 0.13
24 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.28 0.30 0.14 0.19
30 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.34 0.37 0.20 0.25
36 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.39 0.42 0.23 0.29
42 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.43 0.46 0.26 0.34
48 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.47 0.50 0.28 0.37
54 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.51 0.55 0.31 0.42
60 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.54 0.58 0.32 0.44
72 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.59 0.63 0.35 0.51
84 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.62 0.67 0.39 0.54
96 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.68 0.71 0.43 0.63
108 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.71 0.75 0.45 0.65
120 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.75 0.79 0.49 0.68
132 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.77 0.81 0.52 0.72
144 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.78 0.82 0.53 0.74
156 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.80 0.84 0.56 0.75
168 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.81 0.84 0.57 0.80
180 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.83 0.86 0.62 0.82

Percent censored (not yet remarried as of interview) . . . 37.90 34.10 61.30 44.10

– Quantity zero.
NOTE: This table includes all first divorces of women 15–44 years of age in 1995.
 Advance Data No. 323 + May 31, 2001 11
49 percent of black non-Hispanic
100 White non-Hispanic Hispanic Black non-Hispanic divorcees, and 68 percent of Hispanic
divorcees have remarried. The percent
censored (not remarried as of the time
of the interview) for black non-Hispanic
Percent of divorced women who have remarried

80 women in table 8 is much higher than

that of white non-Hispanic or Hispanic
women, reflecting the lower likelihood
of remarriage by the time of interview
60
for black non-Hispanic women.

Second marriage
After 10 years of remarriage,
40 47 percent of remarriages of women
under age 25 years at remarriage have
dissolved, compared with only
34 percent of remarriages to women at
least age 25 years at remarriage (table 9
20
and figure 7). For the first 7 years of
remarriage, the differences are in the
same direction, but they are not
statistically significant.
0 Due to small sample sizes of
0 2 4 6 8 10 12 14
Hispanic and black non-Hispanic
Duration in years
remarriages, differences in the
Figure 6. Probability of remarriage by duration of divorce and woman’s race/ethnicity:
probability of second marriage
United States, 1995 disruption among the three racial/ethnic
subgroups are not statistically significant
At later durations, the curve for Table 8 shows that within 5 years of at any duration of remarriage. However,
Hispanic women converges with that of divorce, 58 percent of white non- table 10 shows that after 10 years of
white non-Hispanic women (and the Hispanic women have remarried, while remarriage, 48 percent of black non-
difference between Hispanic and white 44 percent of Hispanic and 32 percent of Hispanic women’s remarriages,
non-Hispanic women is not statistically black non-Hispanic women have done 39 percent of white non-Hispanic
significant). so. After 10 years of divorce, 79 percent women’s remarriages, and 29 percent of
of white non-Hispanic divorcees,

Table 9. Probability of second marriage disruption by duration of marriage and wife’s age at remarriage: United States, 1995

Age

Proportion of second Less than 25 years

marriages disrupted after— Total 25 years and over

0 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.02 0.03 0.02
12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.05 0.06 0.05
18 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.08 0.09 0.08
24 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.11 0.14 0.10
30 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.13 0.16 0.12
36 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.15 0.18 0.14
42 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.17 0.21 0.16
48 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.20 0.23 0.19
54 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.21 0.24 0.20
60 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.23 0.25 0.22
72 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.26 0.28 0.25
84 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.29 0.34 0.27
96 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.34 0.40 0.30
108 months . . . . . . . . . . . . . . . . . . . . . . . . . . 0.37 0.44 0.32
120 months . . . . . . . . . . . . . . . . . . . . . . . . . . 0.39 0.47 0.34
Percent censored (intact at interview) . . . . . . . . . . 68.50 51.10 75.00

– Quantity zero.
1
Disruption refers to either separation or divorce.
NOTE: This table includes all second marriages of women 15–44 years of age in 1995.
12 Advance Data No. 323 + May 31, 2001
comparable with statistics presented in a
50 Less than 25 years 25 years and over published analysis of the cycle 1 data
(22). The cumulative probability of first
marriage dissolution after 10 years of
marriage was 0.20 in 1973 and 0.33 in
Percent of second marriages disrupted

40 1995. The probability of second

marriage dissolution after 10 years of
marriage was 0.29 in 1973 and 0.39 in
30 1995. Marital dissolution is more likely
now than in the past for both first and
second marriages.
Comparisons of estimates of the
20 probability of marital dissolution by
race/ethnicity are also possible. In the
cycle 1 analysis, there were very few
10 Hispanic women in the sample so
women were grouped by race without
regard to Hispanic origin. Estimates of
marital dissolution were presented for
0 all white women and for all black
0 2 4 6 8 10
women regardless of Hispanic ethnicity.
Duration in years
The cycle 5 estimates of marital
dissolution are recalculated for all white
Figure 7. Probability of second marriage disruption by duration of marriage and wife’s
age at remarriage: United States, 1995
women and all black women to provide
comparability with the cycle 1 analysis.
Comparing the recalculated 1995
estimates with the 1973 estimates, the
Hispanic women’s remarriages have in the sample was too small to be
probability of first marital dissolution
disrupted (table 10 and figure 8). analyzed separately.
after 10 years of marriage was 0.18 for
Although the differences are not
Trends over time white women in 1973 and 0.32 in 1995.
significant, prior research has suggested
The probability of first marital
that the remarriages of black women are The statistics presented in this dissolution was 0.40 for black women in
less stable than those of white women report are based on data from cycle 5 of 1973 and 0.47 in 1995. The probability
(22). McCarthy’s research did not the NSFG, collected in 1995. Cycle 1 of of second marital dissolution after 10
consider Hispanic women separately the NSFG was collected in 1973. Some years of marriage was 0.28 for white
because the number of Hispanic women of the statistics in this report are women in 1973 and 0.39 in 1995. The

Table 10. Probability of second marriage disruption by duration of marriage and wife’s race/ethnicity: United States, 1995

Proportion of second White Black

marriages disrupted1 after— Total non-Hispanic non-Hispanic Hispanic

0 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . – – – –
6 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.02 0.02 0.05 0.02
12 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.05 0.06 0.06 0.05
18 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.08 0.08 0.12 0.06
24 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.11 0.11 0.15 0.06
30 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.13 0.14 0.17 0.07
36 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.15 0.16 0.21 0.08
42 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.17 0.18 0.24 0.10
48 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.20 0.20 0.28 0.14
54 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.21 0.21 0.29 0.16
60 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.23 0.23 0.32 0.17
72 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.26 0.26 0.36 0.19
84 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.29 0.30 0.39 0.22
96 months . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.34 0.34 0.44 0.26
108 months . . . . . . . . . . . . . . . . . . . . . . . . . . 0.37 0.37 0.48 0.26
120 months . . . . . . . . . . . . . . . . . . . . . . . . . . 0.39 0.39 0.48 0.29
Percent censored (intact at interview) . . . . . . . . . . 68.50 67.80 57.90 79.50

– Quantity zero.
1
Disruption refers to either separation or divorce.
NOTE: This table includes all second marriages of women 15–44 years of age in 1995.
 Advance Data No. 323 + May 31, 2001 13
numbers of Asian non-Hispanic women
Black non-Hispanic White non-Hispanic Hispanic were not sufficient for most analyses,
50
the data suggest that the first marriages
of Asian women are more stable than
Percent of second marriages disrupted

those of other groups.

40 Undoubtedly, there are complex
economic, social, and community factors
that affect the differences found in this
30 report. Some of these factors will be
investigated further in a subsequent
report on marriage, divorce, and
20
cohabitation in the United States.

References
10 1. Waite LJ. Does marriage matter?
Demography 32 (4):483–507. 1995.
2. Rogers RG. Marriage, sex, and
mortality. J of Marriage and the
0 Family 57 (2):515–26. 1995.
0 2 4 6 8 10
3. Waite LJ, Gallagher M. The case for
Duration in years marriage: Why married people are
happier, healthier and better off
Figure 8. Probability of second marriage disruption by duration of marriage and wife’s
financially. New York: Doubleday.
race/ethnicity: United States, 1995
2000.
4. Murray JE. Marital protection and
marital selection: Evidence from a
probability of second marital dissolution separations of black non-Hispanic historical-prospective sample of
was 0.29 for black women in 1973 and women are much less likely to make the American men. Demography 37
0.47 in 1995. For both white and black transition to divorce, and the interval (4):511–521. 2000.
women, the probability of marital between divorce and remarriage is 5. Amato PR. The consequences of
dissolution appears to have grown over longer for this group. The data suggest divorce for adults and children. J of
this 22-year period for both first and that the remarriages of black non- Marriage and the Family 62
second marriages. Hispanic women disrupt faster than the (4):1269–87. 2000.
6. Duncan GJ, Hoffman SD. A
remarriages of other women. Some
reconsideration of the economic
Summary and discussion researchers have suggested that these
consequences of marital dissolution.
differences may be related to higher Demography 22 (4):485–97. 1985.
Higher age at first marriage is rates of unemployment, incarceration,
associated with longer marital durations. 7. Smock PJ. The economic costs of
and mortality; lower levels of marital disruption for young women
Age at separation is not significantly educational attainment and earnings; over the past two decades.
related to the probability that a previous experiences as children of Demography 30 (3):353–71. 1993.
separation will result in divorce. Women unmarried or less-educated parents; and 8. Holden KC, Smock PJ. The
at least age 25 years at the time of higher rates of poverty and lack of job economic costs of marital
divorce are less likely to remarry than opportunities in the communities in dissolution: Why do women bear a
women under age 25 years at divorce. which black Americans live (12,18,19). disproportionate cost? Annual
Women younger than 25 years old at Review of Sociology 17 (1):51–78.
Comparisons among women of
remarriage are more likely to experience 1991.
other races are not as consistent. There
a second marital disruption than are 9. Coleman M, Ganong L, Fine M.
is no significant difference between Reinvestigating remarriage: Another
women over age 25 years at remarriage, white non-Hispanic and Hispanic
although this difference only becomes decade of progress. J of Marriage
women in the stability of the first and the Family 62 (4):1288–1307.
statistically significant at later durations marriage. White non-Hispanic women 2000.
of a second marriage. are more likely to make the transition 10. Zill N, Nord CW. Running in place:
In each comparison of racial/ethnic from separation to divorce than Hispanic How American families are faring in
subgroups, the results consistently women and are more likely to remarry a changing economy and an
suggest that the marriages of black than Hispanic women. However, individualistic society. Washington.:
non-Hispanic women are less stable than remarriages among Hispanic people Child Trends, Inc. 1994.
those of white non-Hispanic or Hispanic appear to last longer than the 11. McLanahan S, Sandefur G. Growing
women. The first marriages of black up with a single-parent: What hurts,
remarriages of white non-Hispanic
non-Hispanic women disrupt faster than what helps. Cambridge, MA: Harvard
women, although this final finding is not University Press. 1994.
the first marriages of other women. The statistically significant. Although the
14 Advance Data No. 323 + May 31, 2001
12. Cherlin A. Marriage, divorce and 23. Menken J, Trussell J, Stempel D, 34. Palmore JA, Gardner RW. Measuring
remarriage: Revised and enlarged Babakol O. Proportional hazards life mortality, fertility, and natural
edition. Cambridge, MA: Harvard table models: An illustrative analysis increase: A self-teaching guide to
University Press. 1992. of socio-demographic influences on elementary measures. Honolulu: The
13. Bumpass LL, Sweet JA, Cherlin A. marriage dissolution in the United East-West Center. 1983.
The role of cohabitation in declining States. Demography 18 (2):181–200. 35. Pollard AH, Yusuf F, Pollard GN.
rates of marriage. J of Marriage and 1981. Demographic techniques, 2nd edition.
the Family 53 (4):913–27. 1991. 24. Seltzer JA. Families formed outside Sydney, Australia: Pergamon Press.
14. Oropesa RS, Gorman BK. Ethnicity, of marriage. J of Marriage and the 1981.
immigration, and beliefs about Family 62 (4): pp. 1247–68. 2000. 36. Allison P. Event history analysis:
marriage as a ‘‘tie that binds.’’ 25. Kahn JR, London KA. Premarital sex Regression for longitudinal event
Chapter 10, The ties that bind: and the risk of divorce. J of Marriage data. Sage University Paper series on
Perspectives on marriage and and the Family 53 (4):845–55. 1991. Quantitative Applications in the
cohabitation, Waite LJ, et. al. (ed.). 26. Grady WR. Remarriages of women Social Sciences, 07–046. Beverly
New York: Aldine de Gruyter. 2000. 15–44 years of age whose first Hills and London: Sage Publications.
15. Tucker MB. Marital values and marriages ended in divorce: United 1984.
expectations in context: Results from States, 1976. Advance Data from 37. Sweet JA, Bumpass LL. American
a 21-city survey. Chapter 9, The ties Vital and Health Statistics of the families and households. New York:
that bind: Perspectives on marriage National Center for Health Statistics, Russell Sage. 1987.
and cohabitation, Waite LJ, et. al. 58. 1980. 38. SAS Institute Inc., SAS/STAT User’s
(ed.). New York: Aldine de Gruyter. 27. Abma J, Chandra A, Mosher W, et Guide, Version 6, Fourth Edition,
2000. al. Fertility, family planning, and Volume 2, Cary, NC: SAS Institute
16. Teachman JD, Tedrow LM, Crowder women’s health: New data from the Inc. 1989.
KD. The changing demography of 1995 National Survey of Family
America’s families. J of Marriage Growth. National Center for Health
and the Family 62 (4):1234–46. Statistics. Vital and Health Statistics
2000. 23 (19). 1997.
17. Schoen R. The widening gap 28. Potter FJ, Iannacchione VG, Mosher
between black and white marriage WD, et al. Sample design, sampling
rates: Context and implications. weights, imputation and variance
Chapter 4, The decline in marriage estimation in the 1995 National
among African Americans, Tucker Survey of Family Growth. National
MB and Mitchell-Kernan C (eds.). Center for Health Statistics. Vital and
New York: Russell Sage. 1995. Health Statistics 2 (124). 1988.
18. Wilson WJ. The truly disadvantaged: 29. Kelly JE, Mosher WD, Duffer AP,
The inner city, the underclass, and and Kinsey SH. Plan and operation
public policy. Chicago: University of of the 1995 National Survey of
Chicago Press. 1987. Family Growth. Vital and Health
19. Testa M, Krogh M. The effect of Statistics 1 (36). 1997.
employment on marriage among 30. Clarke SC. Advance report of final
black males in inner-city Chicago. marriage statistics, 1989 and 1990.
Chapter 3, The decline in marriage Monthly Vital Statistics Report 43
among African Americans, Tucker (12), suppl. Hyattsville, Maryland:
MB and Mitchell-Kernan C (eds.). National Center for Health Statistics.
New York: Russell Sage. 1995. 1995.
20. White L, Rogers SJ. Economic 31. Clarke SC. Advance report of final
circumstances and family outcomes: divorce statistics, 1989 and 1990.
A review of the 1990s. J of Marriage Monthly Vital Statistics Report 43
and the Family 62 (4):1035–51. (9), suppl. Hyattsville, Maryland:
2000. National Center for Health Statistics.
21. Bachrach CA, Horn MC. Married 1995.
and unmarried couples, United 32. London KA. Cohabitation, marriage,
States, 1982. Vital and Health marital dissolution, and remarriage:
Statistics, Series 23, No. 15. DHHS United States, 1988. Advance Data
Pub No. (PHS) 87–1991. Public from Vital and Health Statistics 194.
Health Service. Washington. U.S. Hyattsville, Maryland: National
Government Printing Office, July Center for Health Statistics. 1990.
1987. 33. Anderson RN. United States life
22. McCarthy J. A comparison of the tables, 1997. National Vital Statistics
probability of the dissolution of first Reports 47 (28). Hyattsville,
and second marriages. Demography Maryland: National Center for Health
15 (3):345–59. 1978. Statistics. 1999.
 Advance Data No. 323 + May 31, 2001 15

Technical notes Remarriage—Remarriage refers to endured intact to specific duration

the second marriage. Third and higher- values. For further definition of the
Definitions of terms order marriages are not included in the term, see the Methods section.
statistics in this report. Censored—A case is termed
Cohabitation and marital Age at marriage—The wife’s age at ‘‘censored’’ if it is removed from the
status—The categories of past and marriage is calculated as the number of risk of an event before that event has
current cohabitation and marital status months between the date of birth and occurred. For a marriage that is intact at
were derived from two variables in the the date of the marriage, divided by 12, interview, the eventual date of marital
NSFG cycle 5, formal marital status at and truncated to the integer value. disruption is unknown. Therefore, it is
interview and informal marital status at Age at separation—The wife’s age referred to as ‘‘censored by interview.’’
interview. The formal marital status at separation is calculated as the number For a marriage that ended in
variable classifies women according to of months between the date of birth and widowhood, the eventual date of marital
their legal marital status at interview the date of the separation, divided by disruption, had the husband survived, is
(married, widowed, divorced, separated, 12, and truncated to the integer value. unknown. Therefore, it is referred to as
or never married), and the informal Age at divorce—The wife’s age at ‘‘censored by widowhood.’’ For a
marital status variable includes an divorce is calculated as the number of further explanation, see the Methods
additional category, ‘‘not married, but months between the date of birth and section of the text.
living with a partner.’’ Both variables the date of the divorce, divided by 12,
are required to determine the legal and truncated to the integer value. Sample sizes
marital status of cohabiting women. An Age at remarriage—The wife’s age
additional variable, ‘‘ever cohabited The estimates of first marriage
at remarriage is calculated as the duration are based on 6,841 first
outside of marriage,’’ was used to number of months between the date of
determine whether women had ever marriages in the 1995 NSFG. The
birth and the date of the remarriage, estimates of the duration of the
cohabited. divided by 12, and truncated to the
Marital duration—The duration of a separation until the transition to divorce
integer value. are based on 2,386 first marriages that
first marriage is calculated as the Race and Hispanic origin—Women
number of months between the ended in separation. The estimates of
who answered ‘‘yes’’ to the following the duration of divorce until remarriage
beginning of the first marriage and the question were classified as Hispanic:
separation or divorce, or alternatively, are based on 2,033 first divorces. The
‘‘Are you of Hispanic or Spanish estimates of second marriage duration
between the beginning of the marriage origin?’’ Those who answered ‘‘yes’’
and the date of interview, if censored by are based on 1,282 second marriages
were asked ‘‘Are you Puerto Rican, (table I). Table I presents the sample
the interview. The duration of the Cuban, Mexican, or a member of some
second marriage is similarly calculated sizes of subgroups used for estimating
other group?’’ All other women were event probabilities by age and race/
as the number of months between the classified according to race, based on
beginning of the second marriage and ethnicity. Subgroups with less than 100
responses to the following question: cases are not analyzed separately.
the separation or divorce, or between the ‘‘Which of the groups best describes
beginning of the marriage and the date There was one case in the sample
your racial background?’’ The response that reported a later date of divorce
of the interview, if censored by the categories were ‘‘Alaskan native or
interview. from the first marriage than the date of
American Indian,’’ ‘‘Asian or Pacific remarriage, resulting in a negative value
Marital disruption/dissolution— Islander,’’ ‘‘Black’’ and ‘‘White.’’
Disruption and dissolution are used for the duration of divorce until
Women who selected more than one remarriage. This is likely either a
interchangeably and are defined as either race category were asked to select the
separation or divorce. Widowhood is reporting error or a recording error. In
one category that best describes them. addition, 35 cases reported a later date
very rare in the age range for the Alaskan native or American Indian
sample and is not defined as disruption. of final separation from the husband
non-Hispanic women are too small a than the date of divorce. Presumably,
Widowhood removes a marriage from group for reliable estimates from this
the risk of disruption via separation or these are cases where the ex-husband
sample and are omitted from all and ex-wife continued living together
divorce. Because the duration of the analyses by racial/ethnic subgroups.
marriage is unknown if the spouse had after the divorce was finalized, either for
Asian non-Hispanic women are too the sake of the children or for economic
survived, such cases are treated as small a group for reliable estimates for
censored and the date of the spouse’s reasons. All negative duration values
most of the analyses by racial/ethnic were reset to zero, and the time interval
death is used as the end date for subgroups and are omitted as necessary.
calculating duration. until divorce (if separated) or until
Life table analysis—Life table remarriage (if divorced) is equal to zero
Separation—Separation is defined at analysis is a methodology that estimates
the date when the husband and wife in these cases.
probabilities of death at each duration of
stopped living together for the last time. life, adapted for marital stability analysis
Divorce—Divorce is defined at the to estimate probabilities of marital
date the divorce became finalized. events given that the relationship has
16 Advance Data No. 323 + May 31, 2001
Table I. Sample sizes of subgroups of women 15–44 years of age: United States, 1995

Separation
First from first First Second
Subgroup marriage marriage divorce marriage

Total. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6,841 2,386 2,033 1,282

Under 18 years of age . . . . . . . . . . . . . . . . . . . 817 ... ... ...

18–19 years of age . . . . . . . . . . . . . . . . . . . . . 1,532 ... ... ...
20–24 years of age . . . . . . . . . . . . . . . . . . . . . 2,895 ... ... ...
Over 24 years of age . . . . . . . . . . . . . . . . . . . . 1,597 ... ... ...

Under 20 years of age . . . . . . . . . . . . . . . . . . . ... 280 ... ...

20–24 years of age . . . . . . . . . . . . . . . . . . . . . ... 867 ... ...
25–29 years of age . . . . . . . . . . . . . . . . . . . . . ... 657 ... ...
Over 29 years of age . . . . . . . . . . . . . . . . . . . . ... 582 ... ...

Under 25 years of age . . . . . . . . . . . . . . . . . . . ... ... 834 336

25 years of age and over. . . . . . . . . . . . . . . . . . ... ... 1,199 946

Asian non-Hispanic . . . . . . . . . . . . . . . . . . . . . 167 ... ... ...

Black non-Hispanic . . . . . . . . . . . . . . . . . . . . . 1,144 495 352 149
White non-Hispanic . . . . . . . . . . . . . . . . . . . . . 4,452 1,510 1,387 956
Hispanic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1,020 326 247 150

. . . Category not applicable.

NOTE: Age refers to age at the time of the event: age at first marriage, age at separation from first marriage, age at first divorce, age at remarriage.

Sampling errors between subgroups is assessed by to separation or divorce) within the

comparing the boundaries of 95 percent interval. The third column shows the
The statistics presented in this
confidence intervals constructed around number of marriages that are censored
report are based on a sample survey and
each estimate. The confidence interval is (due to interview or widowhood) in the
may, therefore, differ from statistics that
constructed by the formula: p +/- interval. The fourth column shows the
would be obtained if it were possible to
(1.96)(se), where p is the probability effective sample size at the midpoint of
interview all of the 60.2 million women
estimate, and se is the standard error. If the interval. Because the marriages that
that the survey sample represents. The
a positive difference exists between the are censored within an interval are
standard error of an estimate is a
lower limit of the confidence interval assumed to be distributed evenly across
measure of such differences. The
around the larger estimate and the upper the interval, the effective sample size in
software package SUDAAN, version
limit of the confidence interval around a particular interval is equal to the
7.5.4, was used to estimate the standard
the smaller estimate, then the difference effective sample size of the previous
errors used to test for statistical
between the two estimates is statistically interval, minus the failures that occurred
significance. Selected standard errors are
significant at the p< = 0.05 level. in the previous interval, minus one-half
shown in table II.
Like all survey data, the data in this of the censored cases in the previous
In this report, unless otherwise
report may be affected by nonsampling interval, minus one-half of the censored
noted, comparisons between probability
error from such sources as nonresponse, cases in the current interval.
estimates that refer to differences
respondent misreporting due to memory The fifth column of table III shows
between subgroups in the probability of
loss, misclassification of unions or the probability of failure within the
an event (such as marital disruption) are
misreporting of dates, and processing interval and is calculated as the number
statistically significant at the 5-percent
errors. Nonsampling error was of failures divided by the effective
level. This indicates that if the
minimized by stringent quality control sample size. The sixth column shows
difference were merely the result of
procedures incorporated into the survey the cumulative proportion that failed as
random chance and did not reflect a true
design and administration (27). of the beginning of the interval. The
difference in the general population, the
final column in table III shows the
difference would only be observed in Example life table proportion that survived intact until the
less than 5 percent of all possible
Table III presents an example of a beginning of the interval and is simply
samples. Lack of comment about a
life table for the duration of first the complement of the cumulative
specific comparison does not necessarily
marriage. The first column in table III proportion that failed.
indicate a lack of a statistically
lists the duration intervals. In this Columns two through five are
significant difference because not all
example, the intervals are in groups of primarily calculation steps to derive the
possible comparisons are discussed in
12 months, from 0 to 300 (300 months final two columns. In the analysis in this
the text.
= 25 years). The second column shows report, for the sake of brevity, columns
The statistical significance of
the number of marriages that failed (due two through five of each life table are
differences in probability estimates
not shown. Only the cumulative
 Advance Data No. 323 + May 31, 2001 17
Table II. Standard errors for selected probability estimates: United States, 1995

Standard errors for probability estimates after—

Subgroup 3 years 5 years 7 years 10 years 15 years 20 years

First marriage to disruption

Total (tables 3,4) . . . . . . . . . . . . . . . . . . . . 0.0048 0.0062 0.0066 0.0074 0.0087 0.0098
Under 18 years of age at marriage (3). . . . . . . 0.0155 0.0195 0.0205 0.0216 0.0225 0.0227
18–19 years of age at marriage . . . . . . . . . . . 0.0104 0.0124 0.0134 0.0155 0.0152 0.0161
20–24 years of age at marriage . . . . . . . . . . . 0.0061 0.0075 0.0088 0.0105 0.0124 0.0144
Over 24 years of age at marriage. . . . . . . . . . 0.0083 0.0120 0.0140 0.0172 0.0237 0.0414
White non-Hispanic (4) . . . . . . . . . . . . . . . . 0.0058 0.0074 0.0081 0.0087 0.0100 0.0115
Black non-Hispanic . . . . . . . . . . . . . . . . . . 0.0156 0.0178 0.0198 0.0229 0.0242 0.0258
Asian non-Hispanic . . . . . . . . . . . . . . . . . . 0.0202 0.0241 0.0356 0.0420 0.0428 0.0714
Hispanic . . . . . . . . . . . . . . . . . . . . . . . . . 0.0144 0.0159 0.0182 0.0254 0.0313 0.0315

First separation to divorce

Total (tables 5,6) . . . . . . . . . . . . . . . . . . . . 0.0093 0.0069 0.0064 0.0062 0.0060
Under 20 years of age at separation (5). . . . . . 0.0282 0.0233 0.0211 0.0206 0.0202
20–24 years of age at separation . . . . . . . . . . 0.0120 0.0089 0.0077 0.0066 0.0064
Over 24 years of age at separation . . . . . . . . 0.0130 0.0100 0.0095 0.0100 0.0169
White non-Hispanic (6) . . . . . . . . . . . . . . . . 0.0079 0.0054 0.0042 0.0033 0.0033
Black non-Hispanic . . . . . . . . . . . . . . . . . . 0.0315 0.0308 0.0300 0.0322 0.0328
Hispanic . . . . . . . . . . . . . . . . . . . . . . . . . 0.0310 0.0279 0.0295 0.0277 0.0277

First divorce to remarriage

Total (tables 7,8) . . . . . . . . . . . . . . . . . . . . 0.0131 0.0149 0.0149 0.0137 0.0119
Under 25 years of age at divorce (7). . . . . . . . 0.0209 0.0209 0.0209 0.0165 0.0136
25 years of age and over at divorce . . . . . . . . 0.0173 0.0189 0.0202 0.0209 0.0238
White non-Hispanic (8) . . . . . . . . . . . . . . . . 0.0153 0.0165 0.0165 0.0147 0.0130
Black non-Hispanic . . . . . . . . . . . . . . . . . . 0.0283 0.0321 0.0388 0.0445 0.0502
Hispanic . . . . . . . . . . . . . . . . . . . . . . . . . 0.0337 0.0434 0.0481 0.0426 0.0426

Second marriage to disruption

Total (tables 9,10) . . . . . . . . . . . . . . . . . . . 0.0113 0.0138 0.0159 0.0193
Under 25 years of age at remarriage (9) . . . . . 0.0227 0.0254 0.0285 0.0318
25 years of age and over at remarriage . . . . . . 0.0130 0.0168 0.0194 0.0224
White non-Hispanic (10) . . . . . . . . . . . . . . . 0.0125 0.0147 0.0176 0.0212
Black non-Hispanic . . . . . . . . . . . . . . . . . . 0.0440 0.0582 0.0586 0.0635
Hispanic . . . . . . . . . . . . . . . . . . . . . . . . . 0.0254 0.0326 0.0390 0.0499

proportion that failed at the beginning of

each interval is presented and compared
across subgroups. In addition, the
overall percent of cases censored is
presented for each subgroup as a
comparative measure of the overall
likelihood of disruption. In analyses of
the probability of marital disruption,
subgroups with a relatively high percent
censored by interview are less likely to
dissolve (are more likely to have
survived intact until the interview).
18 Advance Data No. 323 + May 31, 2001
Table III. First marriage life table estimates for women 15–44 years old: United States, 1995

Effective Cumulative
Number Number sample Probability proportion Proportion
failed1 censored2 size3 of failure4 failed5 surviving6 as
during during at interval during as of start of start of
interval interval midpoint interval of interval interval
Duration interval x (Fx) (Cx) (ESSx) (PFx) (CFx) (Sx)

0–11 months . . . . . . . . . . . . . . . 1,033,065 1,412,535 36,777,298.5 0.0281 – 1

12–23 months. . . . . . . . . . . . . . . 1,471,780 1,393,281 34,341,325.5 0.0429 0.0281 0.9719
24–35 months. . . . . . . . . . . . . . . 1,605,553 1,354,446 31,495,682.0 0.0510 0.0697 0.9303
36–47 months. . . . . . . . . . . . . . . 1,357,987 1,403,071 28,511,370.5 0.0476 0.1172 0.8828
48–59 months. . . . . . . . . . . . . . . 1,209,263 1,362,065 25,770,815.5 0.0469 0.1592 0.8408
60–71 months. . . . . . . . . . . . . . . 877,368 1,470,283 23,145,378.5 0.0379 0.1987 0.8013
72–83 months. . . . . . . . . . . . . . . 908,833 1,106,335 20,979,701.5 0.0433 0.2290 0.7710
84–95 months. . . . . . . . . . . . . . . 717,709 1,039,565 18,997,918.5 0.0378 0.2624 0.7376
96–107 months . . . . . . . . . . . . . . 543,885 1,124,049 17,198,402.5 0.0316 0.2903 0.7097
108–119 months . . . . . . . . . . . . . 510,398 1,066,175 15,559,405.5 0.0328 0.3127 0.6873
120–131 months . . . . . . . . . . . . . 404,891 1,119,949 13,955,945.5 0.0290 0.3353 0.6647
132–143 months . . . . . . . . . . . . . 406,272 941,474 12,520,343.0 0.0324 0.3546 0.6454
144–155 months . . . . . . . . . . . . . 274,994 1,049,563 11,118,552.5 0.0247 0.3755 0.6245
156–167 months . . . . . . . . . . . . . 236,301 988,623 9,824,465.5 0.0241 0.3910 0.6090
168–179 months . . . . . . . . . . . . . 298,600 878,885 8,654,410.5 0.0345 0.4056 0.5944
180–191 months . . . . . . . . . . . . . 177,560 794,816 7,518,960.0 0.0236 0.4261 0.5739
192–203 months . . . . . . . . . . . . . 217,311 785,294 6,551,345.0 0.0332 0.4397 0.5603
204–215 months . . . . . . . . . . . . . 136,559 775,284 5,553,745.0 0.0246 0.4583 0.5417
216–227 months . . . . . . . . . . . . . 94,765 810,976 4,624,056.0 0.0205 0.4716 0.5284
228–239 months . . . . . . . . . . . . . 109,255 694,968 3,776,319.0 0.0289 0.4824 0.5176
240–251 months . . . . . . . . . . . . . 36,254 674,093 2,982,533.5 0.0122 0.4974 0.5026
252–263 months . . . . . . . . . . . . . 53,307 651,434 2,283,516.0 0.0233 0.5035 0.4965
264–275 months . . . . . . . . . . . . . 46,338 599,026 1,604,979.0 0.0289 0.5151 0.4849
276–287 months . . . . . . . . . . . . . 14,465 407,770 1,055,243.0 0.0137 0.5291 0.4709
288–299 months . . . . . . . . . . . . . 8,880 374,135 649,825.5 0.0137 0.5355 0.4645
300 months or more . . . . . . . . . . . – 453,878 226,939.0 – 0.5419 0.4581

– Quantity zero.
1
Marriages ‘‘fail’’ by separation or divorce.
2
‘‘Censored’’ marriages were still intact at interview.
3
ESSx+1 = ESSx - Fx - ½ (Cx + Cx+1)
4
PFx = Fx / ESSx
5
CFx = 1 - Sx
6
Sx+1 = Sx * (1 - PFx )
20 Advance Data No. 323 + May 31, 2001

Suggested citation Copyright information National Center for Health Statistics

Bramlett MD, Mosher WD. First marriage All material appearing in this report is in the Director
dissolution, divorce, and remarriage: United public domain and may be reproduced or Edward J. Sondik, Ph.D.
States. Advance data from vital and health copied without permission; citation as to
statistics; no. 323. Hyattsville, Maryland: Deputy Director
source, however, is appreciated.
National Center for Health Statistics. 2001. Jack R. Anderson

DEPARTMENT OF
MEDIA MAIL
HEALTH & HUMAN SERVICES POSTAGE & FEES PAID
CDC/NCHS
Centers for Disease Control and Prevention
PERMIT NO. G-284
National Center for Health Statistics
6525 Belcrest Road
Hyattsville, Maryland 20782-2003

OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE, $300

To receive this publication regularly, contact

the National Center for Health Statistics by
calling 301- 458-4636
E-mail: [email protected]
Internet: www.cdc.gov/nchs

DHHS Publication No. (PHS) 2001-1250

01-0384 (5/01)