June 15, 2001
DECLINING SHARE OF CHILDREN LIVED WITH
SINGLE MOTHERS IN THE LATE 1990s
Substantial Differences by Race and
Income
By Allen Dupree and Wendell Primus
Analysis of the Census Bureau's Current Population
Survey shows that between 1995 and 2000, the proportion of children younger than 18 living
with a single mother declined from 19.9 percent to 18.4 percent — a statistically
significant drop of 1.5 percentage points, or 8 percent. This represents a change from the
trend seen ten years earlier. Between 1985 and 1990, the proportion of children living
with a single mother remained at essentially the same level.
In contrast, the proportion of children living with two
married parents (including step-parents) remained essentially unchanged during this
period, at about 70 percent. This differs from the period a decade earlier, when the share
of children living with two married parents declined.
In both time periods, there was a small increase in the
proportion of children living with a cohabiting mother.(1)
Additionally, the share of children living with a single father —while still small
— increased from 3.3 percent to 3.9 percent, continuing the growth seen in the late
1980s.
Children living in lower-income families (families with
income below 200 percent of the official poverty line) are more likely to live with single
mothers and less likely to live with two married parents than are higher-income children.
Among children in lower-income families, there was a significant decline between 1995 and
2000 in the share living with a single mother. There was also an increase in the
proportion of children living with a cohabiting mother and an adult male.
These changes in the living arrangements of children in
lower-income families between 1995 and 2000 reflect a substantive shift from the trends
seen ten years earlier. The proportion of low-income children living with two married
parents fell from 58.1 percent in 1985 to 53.4 percent in 1990. During the same period,
the proportion living with a single mother increased from 32.6 percent to 34.3 percent.
Defining
"Cohabiting Mother"
In this analysis, we use the term "cohabiting
mother" to refer to a living situation in which an unmarried mother lives with a
single adult male who is not related to her and with no other adults. This definition is
consistent with the Census Bureau's concept of the "Persons of the Opposite Sex
Sharing Living Quarters," (POSSLQ) used in its analysis of cohabitation trends over
time.1
In 1995 the Census Bureau began directly asking about the
presence of unmarried partners in the household. Since this direct information was not
collected in the 1985 to 1990 period analyzed in this paper, we use the definition noted
above for both time periods.
It is important to note that the relationships between the
adult male and other members of the household are not known. The male may be the
biological father of one or more of the children in the family. He may have a significant
relationship with the mother or simply be residing in the home.
Also, when making comparisons across income categories, this
analysis utilizes the Census Bureau's standard definition of income, which excludes the
male's income. It is possible that some of these families could have incomes greater than
200 percent of poverty once the male's income is included.
__________________
1 Casper, Lynne, Phillip Cohen and Tavia Simmons,
"How Does POSSLQ Measure Up?" US Bureau of the Census, Population Division
Working Paper No. 36, May, 1999. |
Underlying the overall numbers are substantial
differences across racial and ethnic groups. In the past five years, the proportion of
Black children living with two married parents increased from 34.8 percent to 38.9 percent
— a substantial rise. In addition, an increasing proportion of Black children lived
in homes with a cohabiting mother, while a declining share lived with a single mother.
The proportion of Hispanic children living with a single
mother fell from 24.6 percent in 1995 to 21.3 percent in 2000 — a 13 percent drop.
The proportion of White children living with married parents remained unchanged during
this time, while the proportion living with a single mother decreased slightly.
Following is a more detailed discussion of child living
arrangements as of March of 2000. The paper also discusses the changes in child living
arrangements between 1995 and 2000 and compares these changes with those seen a decade
earlier. In addition, there is an analysis of unusually large changes seen between 1999
and 2000. Finally, the paper offers a brief discussion of trends in child living
arrangements that have been noted by other analyses, including the recently released
Decennial Census data.
Current Child Living Arrangements
The majority of children less than 18 years old — 70
percent — live with two married biological, adoptive, or step-parents.(2)
Another 18.4 percent live with a single mother, while another three percent of children
lived with a cohabiting mother. Four percent of children live with an unmarried father,
while an equal proportion reside with neither parent, either in the care of other
relatives or non-relatives. Less than half of one percent live in some other living
arrangement, typically as the head or spouse of their own family.
A more detailed analysis of children living with a single
mother shows that nearly three-fourths — 73 percent — of these children live
with their mothers and no other adult. Another 22 percent of children living with a single
mother also live with one or more adult relatives. The remaining five percent live in some
other type of home, for example with multiple unrelated adults or a mixture of related and
unrelated adults.
Child living arrangements vary substantially by income. This
analysis separates children into two groups — those living in families with incomes
below 200 percent of poverty, and those living in families with incomes above that level.(3) About half of the children in lower-income families live
with two married parents, compared to 83.6 percent of children in higher-income families.
Poorer children are more than three times as likely as their higher-income counterparts to
live with a single mother, with 32.8 percent of the lower-income children but only 9.5
percent of the higher-income children living with a single mother. Children in
lower-income families also are more likely to live with a cohabiting mother (6.0 percent)
than higher-income children (1.1 percent). Each income group has a similar likelihood of
living with a single father — 4.6 percent of children in the lower-income range and
3.6 percent of children in the upper-income range. Lower-income children, however, are
more likely to live with neither parent — 5.6 percent of lower-income children,
compared with 1.9 percent of upper-income children.
Current living situations also differ greatly across
racial/ethnic categories. White children are somewhat more likely to live with two married
parents than are Hispanic children — 78.2 percent of White children live in such
households, compared with 66.2 percent of Hispanic children.(4)
White children are twice as likely as Black children to live with two married parents
— some 38.9 percent of Black children live in such households. More than 40 percent
of Black children live with a single mother, compared with one-fifth of Hispanic children
and about 12 percent of White children. Black children are somewhat more likely to live in
a cohabiting family (4.2 percent) than White and Hispanic children (2.8 percent and 2.5
percent, respectively). About four percent of children in each racial/ethnic group live
with a single father.
Trends Since 1995
Since 1995, the proportion of children living with married
parents has remained essentially unchanged, but the proportion living with a single mother
has declined. Changes in living arrangements over time, however, differ for children in
lower-income and higher-income families. Among lower-income children, there were no
significant changes in the proportion living with married parents between 1995 and 2000,
while the proportion living with a single mother declined. Among those lower-income
children living with a single mother, however, there was a shift away from sharing a home
with other relatives and towards living with a man who may have been cohabiting with the
mother.
In 1995, 69.5 percent of all children lived with married
parents. This proportion did not change significantly by 2000. There was, however, a
decline in the proportion of children living with a single mother, from 19.9 percent in
1995 to 18.4 percent in 2000. Within the group of children living with a single mother,
there were declines in the proportion of children living with only a single mother, from
14.4 percent of all children in 1995 to 13.5 percent in 2000, and living with a single
mother and some other relative, from 4.5 percent to 4.0 percent. There was an increase in
the proportion who lived with a cohabiting mother, from 2.6 percent to 3.0 percent. The
proportion of children living with a single father, while still a small share overall,
increased from 3.3 percent in 1995 to 3.9 percent in 2000.
These trends differed for children in lower-income and
higher-income families. For example, the chances that a child in a lower-income family
lived with both married parents remained statistically unchanged between 1995 and 2000, at
about 50 percent. In contrast, children in higher-income families became somewhat less
likely to live with two married parents, even though the majority of higher-income
children continued to live in such households. In 1995, 85 percent of higher-income
children lived with two parents; the proportion living with two parents declined to 83.6
percent by 2000.
The two income groups also saw differing trends in the late
1990s with respect to the proportion of children living with a single mother. For
lower-income children, the proportion living with a single mother declined. While children
in higher-income families remain far less likely to live with a single mother, they did
not experience any change in the chances of living in a single-mother family.
There were changes in the types of single-mother families in
which lower-income children lived. Between 1995 and 2000, the proportion of lower-income
children living with other relatives as well as a single mother declined from 6.8 percent
to 5.8 percent. The proportion of children living with cohabiting mothers increased
modestly, from 4.8 percent to 6.0 percent. None of these statistics changed for children
in higher-income families.
During this period, the increase in children living with a
single father occurred among higher-income children. The proportion of higher-income
children who lived with a single father increased from 2.7 percent to 3.6 percent. There
was no change among their lower-income counterparts.
The changes in the late 1990s also varied across
racial/ethnic categories. Between 1995 and 2000, there was no significant change in the
proportion of White or Hispanic children living with two married parents. However, the
proportion of Black children in such families increased from 34.8 percent to 38.9 percent.
While all racial and ethnic groups saw a decrease in the proportion of children living
with a single mother, the greatest decline occurred among Black children (a drop from 47.1
to 43.1 percent). Among Hispanics, the proportion fell from 24.6 percent to 21.3 percent,
while the proportion among White children declined slightly (12.8 percent to 12.0
percent). Black children represented the only group to become more likely to live with a
cohabiting mother over this time period. White children were the only group to experience
a significant increase in the proportion living with a single father during this time
period.
Changing Trends Between Late 1980s and Late 1990s
The trends observed between 1995 and 2000 reflect a
significant change in child living arrangements relative to the 1985-1990 period.(5) The earlier trend of declining proportions of children
living with two married parents leveled off in the later period. While the proportion of
children living with a single mother did not change during the earlier period, it declined
in the later period. The proportion of children living with a cohabiting mother increased
during both periods. The proportion of children living with a single father also
increased. These trends also varied by income and race.
As previously noted, the late 1990s saw no change in the
proportion of children living with two married parents. This was a shift from the trend in
the late 1980s, during which the proportion of children living with married parents fell
from 74.3 percent to 73.1 percent.
The slowdown of the decline in the proportion of children
living with married parents, however, was not universal across income groups. Children in
lower-income families experienced a relatively large decline in the proportion living with
both parents in the late 1980s, from 58.1 percent in 1985 to 53.4 percent in 1990. Between
1995 and 2000, the proportion of lower-income children living with married parents
remained unchanged. The proportion of children in higher-income families who lived with
married parents, on the other hand, declined during both time periods. The decline of 1.4
percentage points in the late 1990s was a continuation of a trend apparent in the late
1980s; the proportion of higher-income children living with two married parents fell from
87.9 percent in 1985 to 86.8 percent in 1990.
The trend in the proportion of children living with a single
mother changed substantially between the two time periods. After increasing slowly for
many years, the proportion of children living with a single mother remained essentially
unchanged during the late 1980s. Between 1995 and 2000, the proportion of children living
with a single mother fell by a small but statistically significant amount. These changes
varied by type of single-mother family. While there was no change in the proportion of
children living with only their mother or with their mother and other relatives in the
late 1980s, both types of single-mother families saw declines in the 1990s. However, the
proportion of children living with a cohabiting mother increased in both time periods.
Again, the changes observed differed across income levels.
Among lower-income children, the proportion of children living with a single mother
increased from 32.6 percent to 34.3 percent during the late 1980s. This increased reversed
in the late 1990s, with the proportion of lower-income children living with a single
mother declining between 1995 and 2000. For higher income children, the proportion living
with a single mother did not change in either time period.
The continuing increase in the proportion of children living
with a cohabiting mother appears to have been driven by changes in lower-income families.
In the late 1980s, the proportion of lower-income children living in families with a
cohabiting mother increased from 2.7 percent to 3.6 percent. In the late 1990s, the
proportion of lower-income children living in these households increased further, from 4.8
percent to 6.0 percent. Among children in higher-income families, there was a significant
increase in the late 1980s in the proportion of children living with a cohabiting mother,
but there was no such increase in the late 1990s.
There has been a continuing trend towards children living
with a single father. In the late 1980s, the proportion of children living with a single
father rose from 2.4 percent to 3.0 percent. In the late 1990s, the proportion of children
living with a single father increased by a similar amount, from 3.3 percent to 3.9
percent. Interestingly, the late 1980s increase was driven by changes among children
living below 200 percent of poverty, while the late 1990s change was driven by the trend
among children living above 200 percent of poverty.
There were notable differences when the trends were examined
by race. Among White children, there were no changes in the proportion living with married
parents in either time period. There was, however, a decline in children living with a
single mother in the late 1990s that did not occur in the late 1980s.
Among Black children, on the other hand, there were
substantial changes between the late 1980s and the late 1990s in the proportion living
with two married parents. This proportion remained essentially unchanged in the late 1980s
but increased substantially in the late 1990s. Black children also saw a significant
decline in the proportion living with a single mother between 1995 and 2000, as well as an
increase in their chance of having a cohabiting mother. Neither of these trends was
apparent in the late 1980s.
Among Hispanic children, trends in the proportion living with
a single mother changed between the late 1980s and the late 1990s. During the late 1980s,
there was essentially no change in the proportion of Hispanic children living with a
single mother, but during the 1990s the proportion of Hispanic children living with a
single mother declined.
Results are
Consistent with Decennial Census Results
The Census Bureau recently released data from the 2000
Decennial Census on family structure. One widely reported finding was that the number of
families with children headed by an unmarried mother increased 25 percent since 1990,
while the similar figure for married couples increased only 6 percent. While these
findings may seem at odds with the results presented in this paper, a closer examination
shows that the two findings do not conflict.
The report on the Decennial Census differs from this paper in
how changes in household structure were measured. The Decennial Census analysis looks at
changes in the number of households in each category, whereas this paper examines changes
in the proportion of children in families of various types. Additionally, the two analyses
differ in terms of the unit of analysis. The Census Bureau's statistics are based on
numbers of families with children, whereas this analysis is based on numbers of children.
While this difference should not have a large impact on the results, it does create a
disparity between the two analyses because children living alone or with non-relatives are
excluded from the Census Bureau's analysis but included in this analysis.
A calculation using the Decennial Census data to determine
the proportion of children in families of various types demonstrates that the differences
between the two analyses are not as great as they may initially appear. Based on this
calculation, the Decennial Census data show 76 percent of households with children were
headed by a married couple in 1990, while about 20 percent were headed by an unmarried
mother. In 2000, the data show 72 percent of households with children were headed by
married couples and 22 percent were headed by an unmarried mother. These results are not
very different from the overall results presented on Table 1. Based on the CPS data, the
proportion of children living with two married parents in 1990 was 73 percent and the
proportion living with an unmarried mother was 21 percent (this figure includes cohabiting
mothers, which is consistent with the Decennial Census figures). These proportions were 70
percent and 21 percent in 2000.(6) In addition, this
analysis is based on children, not families, as is the case with the Census analysis. Once
the data from both sources are presented in terms of living arrangement distributions, the
results appear to be consistent.
Additionally, this analysis pays particular attention to
trend changes that occurred between 1995 and 2000. A comparison the of Decennial Census
from 1990 and 2000, while illustrative, cannot detect changes that are concentrated in
particular years in that 10-year period. Again, Table 1 shows that between 1990 and 1995,
the overall proportion of children living with married parents declined, while the percent
living with a single mother increased. As discussed in this paper, the key trend change
did not occur until after 1995. |
Substantive Changes in 2000
Shifts in child living arrangements tend to be too small to
be statistically significant from year to year. Between 1999 and 2000, however, there were
a number of changes in child living arrangements that were large enough to be
statistically significant. This is noteworthy because changes in family structure tend to
be very gradual and often are not noticeable over a short period of time. For example, the
proportion of children living with two married parents increased from 69.1 percent to 70.1
percent. Similarly, the proportion of children living with a single mother declined from
19.6 percent to 18.4 percent.
Among lower-income children, the share living with a single
mother declined from 35.0 percent in 1999 to 32.8 percent in 2000. Statistically
significant changes also occurred among specific racial/ethnic groups. The proportion of
White children living with married parents increased from 77.3 percent to 78.2 percent,
while the share living with a single mother declined from 13.0 percent to 12.0 percent.
The proportion of Black children living with a cohabiting mother increased from 3.1
percent to 4.2 percent.
Recent Trends Detected in Other Data Sources
Other analyses have reached similar conclusions regarding
child living arrangements in recent years. The Census Bureau recently issued a report
based on the Survey of Income and Program Participation regarding child living
arrangements in 1996. In particular, the report noted that the proportion of children
living with a single parent had stopped rising in the 1990s.(7)
Data from the Urban Institute's National Survey of America's
Families has found similar changes between 1997 and 1999 in child living arrangements.
Nationally, the survey found a significant increase in children living in a two-parent
family coupled with a significant decline in children living with a single parent. While
there was no significant change among children living with two parents when separated by
income (above and below 200 percent of poverty), the decrease in children living with a
single parent was concentrated among lower-income children.(8)
Conclusion
This analysis of CPS data for the late 1990s compared to the
late 1980s indicates that child living arrangements have begun to change in recent years.
The decline in the proportion of children living with two married parents that occurred
during the late 1980s stopped in the late 1990s. At the same time, the proportion of
children living with single mothers decreased.
There were notable variations across racial and ethnic
groups. The increase in children living with married parents was concentrated among Black
children. While all groups experienced a decline in the percent of children living with a
single mother, the decreases were largest between Black and Hispanic children.
Changes in children's living arrangements varied across
income groups as well. In the late 1980s, children in both lower and higher-income
families became less likely to live with married parents and more likely to live with
single mothers. While this continued in the late 1990s for children in higher-income
families, both trends ceased for lower-income children.
Data
The child living arrangements presented in this analysis are
based on calculations using the Census Bureau's March Current Population Survey (CPS). The
Census Bureau releases its own tabulations of living arrangements annually,(9)
but our analysis differs from the Census analysis in two ways. First, the Census analysis
of families excludes unrelated secondary families — groups of two or more
related individuals living in the home of a non-relative. To get a more complete picture
of all children's living arrangements, we have included such individuals in the analysis.
Second, the Census Bureau defines the child's living arrangements based on the marital
status of the head of the household, whereas this analysis defines a child's living
arrangements based on the presence and marital status of the parent of that child. This
makes a difference in a multi-generational household. For example, if the home contains a
married couple, their unmarried daughter, and her son, Census considers the grandson of
the household to live in a married couple family. Our analysis classifies him as living
with a single mother and other relatives.
The primary advantage of the CPS is that it provides annual
data over an extended time period, which is ideal for analyzing trends. However, there are
some limitations that have affected this analysis. First, the Census Bureau made
substantial changes in its data processing methodology in 1994 that affected the family
relationships shown in the data.(10) This change makes it
difficult to compare statistics derived from data collected before 1994 to statistics
derived from data collected after the change in methodology. However, it does not affect
the comparison of trends before and after the new method was implemented.
In addition, because the data show the relationship of the
individuals in the household only to the designated head of household, it is not possible
to determine the relationship between children and members of the household other than the
household head. Thus, we are unable to clearly differentiate between children living with
two married biological or adoptive parents as opposed to those living with a step-parent.
In the case of a single mother living with an unrelated man, it is impossible to determine
whether or not the man is the biological father of the children.
There are other issues related to the determination of
household income. Income data collected in the March CPS are based on the current
household members' income in the previous calendar year. If a portion of the household's
income in the previous year was obtained by a family member who no longer resides in the
household as of March, that income will not be included as household income in the survey.
This could put the household in an income group that does not reflect the household's
actual income in the previous year. In addition, no income data are collected for children
less than 15 years of age who do not live with relatives. Therefore, children under 15 not
living with any relatives are included in the overall proportions of children in each type
of living arrangement, but excluded from the tables of lower- and higher-income children.
This means that the figures that are disaggregated by income are not completely comparable
to the overall numbers.
End Notes:
1. The CPS data do not identify
non-marital relationships in all years examined. For this analysis, we use the term
"cohabiting" to refer to a single mother living with one unmarried adult male.
The male may in fact be the biological father of one or more of the children in the
household.
2. The Current Population Survey
does not distinguish between step-families and other types of families. See the Data
section for further discussion of this limitation.
3. This analysis uses the Census
Bureau's definition of family income which does not include the unrelated male's income.
4. In this paper, "White"
means White, non-Hispanic, and "Black" means Black, non-Hispanic.
5. Due to a change in the Census
Bureau's data processing methods in 1994, we have not analyzed the trend between 1990 and
1995.
6. The CPS comparison between 1990
and 2000 should be interpreted with caution. As discussed in the data section, changes in
data processing made in 1994 make comparing absolute levels somewhat suspect, although it
would not affect comparisons of trends as done in this paper.
7. Fields, Jason, "Living
Arrangements of Children: Fall 1996," Current Populations Reports, P70-74. U.S.
Census Bureau, Washington, DC, 2001.
8. Vandivere, Sharon, Kristen
Moore, and Martha Zaslow, "Children's Living Environment: Findings from the National
Survey of America's Families," Urban Institute, October 2000.
9. See the P20 series of Census
publications.
10. Saluter, Arlene, "Marital
Status and Living Arrangements: March 1994," Current Populations Report, P20-484,
Census Bureau, Washington, DC, 1996.
