Journal of Interdisciplinary History, xxxviii:1 (Summer, 2007), 33–64.

SINGLE PARENTHOOD AND CHILDHOOD OUTCOMES

Howard Bodenhorn
Single Parenthood and Childhood Outcomes in
the Mid-Nineteenth-Century Urban South An in-
creasing volume of research documents the rising incidence of
single parenthood and its consequences for children. Issues sur-
rounding the consequences of single motherhood have special
salience for black Americans. According to Fields, about 16 per-
cent of white children and 48 percent of black children live with
a single mother, and the economic costs to children raised in sin-
gle-mother households are substantial. More than one-half of sin-
gle-mother households fall under the ofªcial poverty line and
three-ªfths receive some form of public assistance. Moreover,
children raised in mother-only households are more likely to un-
derachieve academically, to drop out of school, to become single
parents themselves, to have lower labor-market attachment, and to
engage in criminal activity as young adults than children raised in
two-parent households. Thus, racial differences in family structure
have potentially large implications for racial differences in child
welfare and individual economic mobility in later life.1

Similar, if less pronounced, racial differences in family struc-
ture are well documented for the late nineteenth and early twenti-
eth centuries. Du Bois, Frazier and, more recently, Sacerdote

Howard Bodenhorn is Professor of Economics, Lafayette College. He is the author of State
Banking in Early America: A New Economic History (New York, 2003); “The Mulatto Advan-
tage: The Biological Consequences of Complexion in Rural Antebellum Virginia,” Journal of
Interdisciplinary History, XXXIII (2002), 21–46.

The author thanks Susan Averett, Robert Margo, Carolyn Moehling, John Murray,
Greg Price, and participants at the Cliometric Society session at the 2006 ASSA annual meet-
ings for valuable comments, and Pam Bodenhorn and Veronica Hart for valuable research as-
sistance. He is also grateful to the National Science Foundation (SES-0109165) and the
Earhart Foundation for ªnancial support.

© 2007 by the Massachusetts Institute of Technology and The Journal of Interdisciplinary
History, Inc.

1
Jason Fields, “Children’s Living Arrangements and Characteristics: March 2002,” Current
Population Reports, U.S. Census Bureau, Study P20-547; Sara McLanahan and Larry
Bumpass, “Intergenerational Consequences of Family Disruption,” American Journal of Sociol-
ogy, XCIV (1988), 130–152; McLanahan and Gary Sandefur, Growing Up with a Single Parent:
What Hurts, What Helps (Cambridge, Mass., 1994); Percy Gamble Kammerer, The Unmarried
Mother: A Study of Five Hundred Cases (Montclair, N.J., 1969); Carolyn M. Moehling, “Family
Structure, School Attendance, and Child Labor in the American South in 1900 and 1910,”
Explorations in Economic History, XLI (2004), 73–100.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

34 | HOWARD BODENHORN

contend that slavery damaged the black family and gave birth to its
modern dysfunctions including single motherhood. The contro-
versy surrounding Moynihan’s now infamous report, which re-
peated Du Bois and Frazier’s mostly unproved assertions,
prompted new historical studies, which found that whether en-
slaved or free, most black children lived in traditional two-parent
families, though free black children were about twice as likely as
white children to live in mother-only households. Mathis and
Krech argue that comparing black households to the nuclear, two-
parent, white norm is hegemonic, failing to recognize the cultural
distinctiveness of extended kinship networks among blacks. Nev-
ertheless, even Krech notes that adults provide more education
and superior employment opportunities for their own children
than for nieces, nephews, and other less direct kin.2

This study documents the incidence of single parenthood in
the urban South in the mid-nineteenth century. Consistent with
the ªndings of other studies of early black households, African
American children were about twice as likely as white children to
reside in mother-only households. The second, and novel, aspect
of this study is that it documents the consequences for children re-
siding in a household with one or neither parent. Although pre-
vious studies, including those by Gutman, Morgan et al., and
Ruggles document differences in household structure between
free blacks and whites in the period before general emancipation,
they do not investigate how differences in household structure

2 William E. Burghardt Du Bois, The American Negro Family (Cambridge, Mass., 1970);
E. Franklin Frazier, The Negro Family in the United States (New York, 1951). Bruce Sacerdote,
“Slavery and the Intergenerational Transmission of Human Capital,” Review of Economics and
Statistics, LXXXVII (2005), 217–234, also shows that not until two generations after emanci-
pation did the educational outcomes of the descendants of slaves resemble the outcomes of
descendants of antebellum free blacks. Daniel P. Moynihan, The Negro Family: A Case for Na-
tional Action (Washington, D.C., 1965); Frank F. Furstenberg, Jr., Theodore Hershberg, and
John Modell, “The Origins of the Female-Headed Black Family: The Impact of the Urban
Experience,” Journal of Interdisciplinary History, VI (1975), 211–233; Herbert G. Gutman, “Per-
sistent Myths about the Afro-American Family,” Ibid., XVI (1975), 181–210; Linda Gordon
and McLanahan, “Single Parenthood in 1900,” Journal of Family History, XVI (1991), 97–116;
Steven Ruggles, “The Origins of African-American Family Structure,” American Sociological
Review, LIX (1994), 136–151; Arthur Mathis, “Contrasting Approaches to the Study of Black
Families,” Journal of Marriage and the Family, XL (1978), 667–676; Shepard Krech III, “Black
Family Organization in the Nineteenth Century: An Ethnological Perspective,” Journal of
Interdisciplinary History, XII (1982), 429–452.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 35

inºuenced several childhood outcomes, including school atten-
dance and labor-market participation.3

A new sample of white and African American children drawn
from the manuscripts of the 1860 census provides important in-
sights into the consequences of being raised in a single-parent
household. First, children younger than nine years old and teenage
youths between fourteen and sixteen years of age residing in
mother-only households were less likely to attend school than
their counterparts living in two-parent households. During the
prime school attendance ages (nine through thirteen years), single
motherhood had modest negative effects on school attendance for
either race. Black children were less likely than whites to attend
school during these years, but their low attendance rates were
more a function of their race than the parental structure of their
households. Children living in mother-only houses started school
later and quit school earlier than children and youths from two-
parent households.

Second, single motherhood was associated with an increased
likelihood that white, but not black, youths between the ages of
fourteen and sixteen would join the labor force in some capacity.
As modern studies attest, the consequences of single motherhood
on social and economic mobility are potentially large; they were
so in the mid-nineteenth century as well. Only longitudinal data
can fully capture the lifetime consequences of growing up in a sin-
gle-parent household, but cross-sectional data from the mid-
nineteenth century do not provide much reason for believing that
the consequences of single motherhood were less profound then
than now.

the historical incidence of single parenthood and its con-
sequences for children A multitude of studies have docu-
mented the increasing incidence of single motherhood in the
United States during the past half-century. In 1940, about 5 per-
cent of white and 15 percent of black households were headed by
single mothers. During the subsequent half-century, the incidence
of single motherhood tripled for black and white households; by

3 Gutman, “Persistent Myths,” 181–210; S. Philip Morgan et al., “Racial Differences in
Household and Family Structure at the Turn of the Century,” American Journal of Sociology,
XCVII (1993), 799–828; Ruggles, “Origins,” 136–151.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

36 | HOWARD BODENHORN

the mid-1980s, nearly half of all black households were headed
by a single mother and about 61 percent of black children did not
live with two biological parents.

This section provides a brief discussion of persistent historical
racial differences in single motherhood and its likely causes. As
Morgan et al. note, the historical evidence is a reminder that racial
differences in family structures are not new. Indeed, differences in
household structure are seemingly rooted in long-standing differ-
ences in family formation processes, many of which are inherited
from the past. Since the social, economic, and cultural history of
African Americans differs from that of whites, it is not surprising
that these differences are reºected in household structures.4

Tocqueville may not have been the ªrst to argue that slavery
destroyed the black family, but his early statement of that hypothe-
sis became the conventional wisdom for generations of historians
and sociologists. Even as the common wisdom among historians
was gradually evolving in the 1970s, the cliometric revolution
overturned it. Fogel and Engerman argued that slaveholders faced
numerous incentives to encourage stable, two-parent families.
From a sample of slave plantations, Gutman reports two distinct
patterns among slave families. One was the traditional male-
headed, two-parent household. The other was the single-mother
household in which an adult male had never resided, though
neither the death nor the sale of a once-resident husband and
father was the reason for the broken home. Gutman could offer no
explanation of why slaveholders would have devised such a di-
chotomous system. He concluded that the household structures
were chosen by slaves, not imposed by slaveholders. Crawford,
who reconstructed patterns of the slave family from the Works
Progress Administration’s (wpa) slave narrative project, also found
evidence of dual family structures. About two-thirds of slave chil-
dren were raised by, or had regular contact with, both biological
parents. The other third were raised in predominantly single-
mother households. Sale disrupted slave marriages, but it does not
follow that the disruption or the threat of disruption diminished

4
Irwin Garªnkel and McLanahan, Single Mothers and Their Children: A New American
Dilemma (Washington, D.C., 1986); David T. Ellwood and Jonathan Crane, “Family Change
among Black Americans: What Do We Know?” Journal of Economic Perspectives, IV (1990), 65–
84; Morgan et al., “Racial Differences.”

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 37

black appreciation for the virtues of two-parent, nuclear house-
holds.5

How did African American family structures established in
slavery make the transition to freedom during the antebellum era?
The few extant studies of the free black families in the period be-
fore general emancipation show that African Americans over-
whelmingly adopted the traditional, nuclear, two-parent structure.
Gutman, relying on a small sample of free blacks residing in Buf-
falo, New York, in 1850, discovered that about 90 percent of
households had an adult male present. It is not clear, however, that
Buffalo’s small free black population in 1850 was representative of
free black communities elsewhere in the United States, especially
those in the Old South or Gulf Coast cities.6

Furstenberg et al.’s investigation of the incidence of single
motherhood in the relatively large free black population in ante-
bellum Philadelphia disclosed that 22.5 percent of black house-
holds were headed by single females—nearly twice the rate for
households headed by native-born whites and Irish immigrants,
and nearly seven times the rate for German immigrants. Using a
nationally representative sample drawn from the 1850 Integrated
Public Use Microdata Series (ipums), Ruggles estimated that about
53 percent of black children fourteen years of age or younger re-
sided with both parents, compared to 83 percent of white chil-
dren. That a host of studies document a relatively constant share
(20 to 25 percent) of mother-only black households between 1870
and 1940 suggests a persistence that is unlikely to be explained
solely by slavery or its legacy. Matrifocal households, to use
Frazier’s term, have been much more common in the black than
the white community for about two centuries, in both slavery and
freedom.7

5 Alexis de Tocqueville (trans. Harvey C. Mansªeld and Delba Winthrop), Democracy in
America (Chicago, 2000; orig. pub. 1835); Du Bois, American Negro Family; Frazier, Negro Fam-
ily; Winthrop D. Jordan, White over Black: American Attitudes toward the Negro, 1550–1812 (Cha-
pel Hill, 1968); John W. Blasingame, The Slave Community: Plantation Life in the Antebellum
South (New York, 1979); Robert W. Fogel and Stanley L. Engerman, Time on the Cross: The
Economics of American Negro Slavery (New York, 1974), 2v.; Stephen Crawford, “The Slave
Family: A View from the Slave Narratives,” in Claudia Goldin and Hugh Rockoff (eds.), Stra-
tegic Factors in Nineteenth Century American Economic History: A Volume to Honor Robert W. Fogel
(Chicago, 1992), 331–350; Gutman, “Persistent Myths.”
6 Gutman, “Persistent Myths.”
7 Furstenberg et al., “Origins”; Ruggles, “Origins.” Du Bois, American Negro Family, re-
ports that 20% of black children born in Washington, D.C., between 1879 and 1907 had un-

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

38 | HOWARD BODENHORN

The question of why single parenthood was more common
among blacks than whites remains unresolved. Unlike in modern
experience, divorce was rare in the mid-nineteenth century, but
simple abandonment may have served the same purpose. Consis-
tent with modern experience, however, Frazier argues that much
black single motherhood was by choice, advancing a prescient
version of Becker’s theory of family. Frazier argues that the em-
ployment opportunities available to black women, mostly in do-
mestic service, and the low-earning capacity of black men as com-
mon laborers, encouraged black women to establish mother-only
households. The traditional marriage contract—in which wives
keep house and raise children in return for husbands’ provision of
subsistence—was unattractive to many women because partner
complementarities were low. Lebsock’s careful study of women in
antebellum Petersburg, Virginia, provides powerful support of
Frazier’s thesis, ªnding that women, black and white alike, with
skills and resources, or unpromising marriage prospects, were
likely to eschew marriage.8

Other common explanations focus on unbalanced sex ratios
and racial differences in early mortality. Although Morgan et al.
and Ruggles turned up little evidence that excess early mortality
among black men can explain racial differentials in single mother-
hood, the incidence and consequences of racial differences in
mortality remain an open question. Explanations relying on un-
balanced sex ratios date to at least Du Bois, who noted a national
sex imbalance of 1,013 black women for every 1,000 black men in
1900. But the national ªgure disguised sharper imbalances in ur-
ban areas. In fourteen of the ªfteen cities with the largest black
populations, black women outnumbered black men by an average

wed mothers. Gordon and McLanahan, “Single Parenthood,” ªnd that in 1900, 87% of white
children, but only 59% of black children, lived with both parents. Morgan et al., “Racial Dif-
ferences,” ªnd that in 1910, the incidence of black single motherhood was about 1.5 times
that of whites. In some cities, however, the incidence among blacks was about four times that
of whites. Kammerer, Unmarried Mother, reports that in 1911, out-of-wedlock births ac-
counted for 2.1 percent of all births among whites, but 22.1 percent among blacks in Wash-
ington, D.C. Frazier, Negro Family, reports that the black illegitimacy rate in the rural South
during the 1930s was 15.4%, and that in the urban South in 1940, 31.1% of black households
were headed by a single female.
8 Frazier, American Negro Family; Gary Becker, A Treatise on the Family (Cambridge, Mass.,
1981); Suzanne Lebsock, The Free Women of Petersburg: Status and Culture in a Southern Town,
1784–1860 (New York, 1984).

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 39

of 1,180 women to 1,000 men. No comparable disparity was
identiªed in any city for whites.9

Sex imbalances continue to be offered as explanations of the
modern experience, often in conjunction with Becker’s partner-
complementarities hypothesis. These explanations are based on
data revealing that marriage rates vary positively with sex ratios
and the supply of eligible men who enjoy stable earnings pros-
pects. The models of Willis and Neal, in which both factors play a
role, show that the poorest women remain childless in equilib-
rium. Women in other segments of the income distribution may
have a choice between marriage and single motherhood, but small
shocks to the marriage market, either in earnings capacities or in
the supply of attractive partners, can lead to substantial differences
in their likelihood of becoming single mothers. Evidence pre-
sented below is consistent with this interpretation. Sharply unbal-
anced sex ratios, especially in the prime childbearing years, in the
urban antebellum South were associated with high rates of single
motherhood, probably by choice.10

data and methodology The effects of family structure on
the incidence of childhood poverty, school attendance, and labor-
market participation are studied using data drawn from the popu-
lation manuscripts of the 1860 census for Baltimore, Maryland,

9 Furstenberg et al, “Origins”; Morgan et al., “Racial Differences”; Ruggles, “Origins”;
Du Bois, American Negro Family. Arnstein Aassve, “Economic Resources and Single Mother-
hood: Incidence and Resolution of Premarital Childbearing among Young American
Women,” Max Planck Institute for Demographic Research, working paper 2000-015
(Rostock, Germany, 2000), also notes a fourth potential cause of single motherhood, social-
welfare beneªts. Given the limited public assistance offered to the poor in early America, in-
cluding the poorhouse, social assistance is unlikely to have contributed meaningfully to out-
of-wedlock childbearing in antebellum America, but it is a question worthy of further study.
Nevertheless, evidence from the Baltimore almshouse between 1833 and 1843 shows a rising
number of claims by adult black women, but a relatively constant number of claims by black
children. Joseph Ferrie, “The Rich and the Dead: Socioeconomic Status and Mortality in the
United States, 1850–1860,” in Dora Costa (ed.), Health and Labor Force Participation over the Life
Cycle: Evidence from the Past (Chicago, 2003), 11–50, challenges the hypothesis that premature
mortality was largely random, arguing that excess mortality was higher among the poor.
10 William Julius Wilson, The Truly Disadvantaged (Chicago, 1987); Wilson, “The Woes of
the Inner-City African American Father,” in Obie Clayton, Ronald B. Mincy, and David
Blankenhorn (eds.), Black Fathers in Contemporary American Society: Strengths, Weaknesses, and
Strategies for Change (New York, 2003), 9–29; Becker, Treatise; Robert J. Willis, “A Theory of
Out-of-Wedlock Childbearing,” Journal of Political Economy, CVII (1999), S33–64; Derek
Neal, “The Economics of Family Structure,” National Bureau of Economic Research work-
ing paper 8519 (2001).

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

40 | HOWARD BODENHORN

and New Orleans, Louisiana (details concerning the construction
of the sample are discussed in the appendix). The choice of these
two cities is driven by three factors. First, from a pragmatic stand-
point, both cities had large and well-established free black com-
munities by the late antebellum era. The size and strength of New
Orleans’ black community is well documented, and Whitman ex-
plores the emergence of Baltimore’s black community early in the
nineteenth century. Second, since this study is part of a larger proj-
ect considering the economic condition of free blacks in the ante-
bellum South, the focus in this article is on southern cities. Third,
by the middle of the nineteenth century, nearly every southern
state, except Maryland and Louisiana, prohibited the education of
free blacks. Maryland excluded black children from public educa-
tion but allowed them to attend private and parochial institutions.
Hence, churches and concerned blacks and whites began charity
schools and academies of their own. The more well-to-do black
families could send their children to any private academy or tutor
willing to accept them. In antebellum New Orleans, free black
children could attend public schools or a number of private and
parochial schools and academies that served the city’s large black
community.11

11 Recent studies of antebellum New Orleans include Paul F. Lachance, “The Formation
of a Three-Caste Society: Evidence from Wills in Antebellum New Orleans,” Social Science
History, XVIII (1994), 211–242; idem, “The Limits of Privilege: Where Free People of Colour
Stood in the Hierarchy of Wealth in Antebellum New Orleans,” Slavery and Abolition, XVII
(1996), 65–84. T. Stephen Whitman, The Price of Freedom: Slavery and Manumission in Baltimore
and Early National Maryland (Lexington, Ky., 1997). Carter G. Woodson, The Education of the
Negro prior to 1861 (New York, 1968), provides a summary of black educational experiences
and opportunities across the South. For more information on Baltimore’s public school sys-
tem, which commenced in 1829, see William H. Shannon, “Public Education in Maryland
(1825–1868) with Special Emphasis upon the 1860s,” unpub. Ph.D. diss. (University of Mary-
land, College Park,1964); Tina H. Sheller, “The Origins of Public Education in Baltimore,
1825–1829,” History of Education Quarterly, XXII (1982), 23–42. For a discussion of New
Orleans’ antebellum school system and educational reforms elsewhere in the South, see Jona-
than Daniel Wells, The Origins of the Southern Middle Class, 1800–1861 (Chapel Hill, 2004),
133–153. Advertisements published in Baltimore during the early 1850s suggest that even the
middling sorts could afford to send children to private academies. The Academy of Visitation
charged $40 per annum, payable in quarterly installments, for elementary education. Children younger than ten years old could attend for $20; those between ten and twelve paid $25 (Balti- more American, 3 Aug. 1850). The Baltimore Collegiate Institute for Young Ladies charged $20
per annum for elementary education and between $40 and $60 per annum for more advanced
instruction (Baltimore American, 8 Sept. 1851). John Hope Franklin and Loren Schweninger, In
Search of the Promised Land: A Slave Family in the Old South (New York, 2006), report that even
though private schools were ofªcially prohibited in antebellum Nashville, Tennessee, they

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 41

One shortcoming of the 1860 census is that its enumerators
asked only whether an individual had received any instruction
during the past year at a public or private school or with a tutor.
Children and youths whose educations were limited to Sunday
school were not recorded as having been “at school.” The stric-
tures of this question permit researchers no way to determine
whether a child received instruction regularly, intermittently, or
even just once. Moehling and Margo came to different conclu-
sions in their interpretation of a similar question in later censuses.
A prudent interpretation of responses in the 1860 census is occa-
sional attendance. Because a Maryland legislative committee in
1860 reported that less than one-half of eligible children attended
schooling regularly, the 1860 census data likely provide an upper
bound estimate to regular school attendance. The exclusion of
Sunday school instruction, however, may underestimate black
schooling. Although Sunday schools focused on religious instruc-
tion, churches often provided blacks with remedial academic
instruction.12

One advantage of using the 1860 census in a study of school
attendance is that it permits control for the effect of household re-
sources. Unlike earlier and later censuses (1870 excepted), the
1860 census collected information about each household’s hold-
ings of real and personal estate. Census enumerators in 1860 also
collected information about the occupations of household heads,
as well as the occupations of other household members aged
ªfteen years and older. Additional information includes the age,
sex, literacy, and place of birth of each household member. A
household’s decision to send a child to school for some part of the
academic year depended on a number of factors, and, relative to
other nineteenth-century censuses, the 1860 census provides a
broad panel of controls, most importantly household resources.

Wealth, literacy, household size, race, and immigrant status
are all believed to have inºuenced the decision to send a child to
school. Less information is available about the supply of schooling
services. It is unlikely that every child of a given race in a given
city had equal access to school. A series of census-ward dummies

operated with the tacit approval of local whites. Nevertheless, private black schools were
forced to close or relocate when whites protested.
12 Moehling, “Family Structure”; Robert A. Margo, Race and Schooling in the South, 1850–
1950 (Chicago, 1990); Shannon, Public Education.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

42 | HOWARD BODENHORN

are included to control for local provision of schools, but dummy
variables cannot fully capture local differences in availability and
accessibility.

Panel A of Table 1 provides a comparison of family structures
for black and white children. Black children were about twice as
likely as whites to live in a single-mother or female-headed house-
hold. Black children were also about three times as likely to live
with neither parent; this arrangement was much more common
among blacks than whites. The incidence of single fatherhood was
twice as likely for whites as blacks.

What explains the higher incidence of single motherhood
among blacks? The most likely explanation is Du Bois’ contention
that sex imbalances in southern cities led to higher rates of single
motherhood among black women who simply chose to have chil-
dren out of wedlock.13

Table 2 reports sex ratios for whites and African Americans in
Baltimore and New Orleans in 1860. In the prime child-bearing
years (twenties and thirties), white men outnumbered white
women. Among free blacks, however, women outnumbered men
by a large margin. The likelihood that black women entering the
marriage market in their twenties would ªnd mates was low com-
pared to that of most other populations. There were 1,675 free
black women for every 1,000 free black men in Orleans Parish,
Louisiana, and 1,709 free black women for every 1,000 free black
men in Baltimore. The sex imbalances were less pronounced for
blacks in their thirties, but this age cohort comprised 1.5 times as
many free black women as men. Assuming that the African Amer-
ican marriage market transcended legal status and that marriages
between slaves and free blacks were permissible and viable, the sex
imbalances become more rather than less pronounced.

The evidence is consistent with theoretical and empirical
studies in which marriage rates vary negatively with the female-
male sex ratio and the supply of men with stable earnings pros-
pects. When women have their own resources or earnings and
when they outnumber men, a large segment of the female popula-
tion invariably favors out-of-wedlock childbearing.14

13 Du Bois, American Negro Family.
14 William Darity, Jr., and Samuel L. Myers, Jr., “Changes in Black Family Structure:
Implications for Welfare Dependency,” American Economic Review, LXXIII (1983), 59–64;
Neal “Economics of Family Structure”; Willis, “Theory.”

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 43

Table 1 Family Status of Resident Children and Youth by Race, Baltimore
and New Orleans (Percent of Children Six to Sixteen Living in
Household by Family Status)

child’s race (%)

white

black

a. household structures

Single mother
Female head
Single father
Neither parent
Traditional family
Two-parent family

b. school attendance during the past year

Boys 6–8 years
Boys 9–13 years
Boys 14–16 years
Girls 6–8 years
Girls 9–13 years
Girls 14–16 years

c. labor-force participation

Boys 6–8 years
Boys 9–13 years
Boys 14–16 years
Girls 6–8 years
Girls 9–13 years
Girls 14–16 years
total observations

7.9
9.5
10.1
2.4
75.9
77.0

59.1
81.5
48.5
62.6
82.9
39.6

0.0
0.7
23.8
0.0
0.3
6.4
6,133

17.1
22.7
5.9
8.9
63.9
63.6

25.1
39.7
12.8
26.3
34.2
11.4

0.0
1.7
28.1
0.0
1.5
13.3
4,561

notes
count for sampling procedures.

See Appendix for deªnitions of household structures. All statistics weighted to ac-

Panel B of Table 1 reports school attendance rates by age and
sex. Racial disparities in school attendance are even more pro-
nounced than differences in household structure. Black children
and youths of both sexes attended school at one-fourth to one-half
the rate of whites. The age groupings reºect generally accepted
mid-nineteenth-century educational norms. Vinovskis reported
general resistance to sending children under nine years of age to
school, and little attendance prior to age six. The “prime” school-
age years at mid-century were nine to twelve or thirteen years,
after which attendance dropped off dramatically. Most Americans
believed that schooling during the formative pre-adolescent years

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

44 | HOWARD BODENHORN

Table 2

Sex Ratios by Race and Age for Baltimore and New Orleans in 1860
(Females per 1000 Males)

age group

whites

free blacks

all blacks

baltimore, maryland

15–19
20–29
30–39
40–49
50–59

1,278
1,186
994
983
1,080

1,972
1,709
1,432
1,241
1,901

2,026
1,726
1,493
1,305
1,959

new orleans, louisiana

age group

whites

free blacks

all blacks

15–19
20–29
30–39
40–49
50–59

1,252
974
891
630
772

1,267
1,675
1,635
1,505
1,804

1,483
1,496
1,529
1,570
1,633

notes All Blacks category includes free blacks and slaves.
source U.S. Bureau of the Census (1864).

was beneªcial; enough common schools operated in most places
to accommodate (white) children in this age group.15

The pattern of school attendance for both blacks and whites is
consistent with Vinovskis’ contention, although attendance rates
for six- to eight-year-olds is higher than a “general” resistance to
sending young children to school would have warranted. Never-
theless, school attendance among nine- to thirteen-year-olds rose
dramatically. Although school attendance rates by free black chil-
dren was low relative to those of whites, the indication that 35 to
40 percent of black children at the prime schooling age attended
for some part of the year when public facilities were generally
unavailable is a testament to the “improving spirit” among these
cities’ free black inhabitants. Another notable feature is that girls
attended school at approximately the same rate as boys, regardless
of race.

Finally, Panel C of Table 1 reports statistics on child labor-
force participation. Given the tendency of census enumerators in

15 Maris A. Vinovskis, “Quantiªcation and the Analysis of American Antebellum Educa-
tion,” Journal of Interdisciplinary History, XIII (1983), 761–786.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 45

1860 to underreport occupational
information for individuals
other than household heads, these rates are probably best inter-
preted as lower bounds, reºecting regular paid employment rather
than intermittent or occasional labor-force participation. Although
the census enumerators were instructed to record occupations or
apprenticeships only for those older than ªfteen years of age, a few
enumerators recorded occupations for children as young as ten.
Most of the children with listed occupations, however, were four-
teen to sixteen years old. About 5 percent of fourteen-year-old
males, 20 percent of ªfteen-year-olds, and 30 percent of sixteen-
year-old males had recorded occupations. Labor-force participa-
tion rates among girls were considerably lower than those for
boys.

Empirical Approach The empirical strategy adopted in this
article follows Moehling, who extended the method originally
developed by Margo. The analysis empirically estimates the effects
of three alternative household structures—single motherhood,
single fatherhood, and children residing with neither parent—
against the alternative of the intact traditional, two-parent house-
hold. Unlike the data in other studies of the modern experience,
the data in this one do not allow for a consideration of the effects
of stepparents or the causes of family disruption on childhood out-
comes.16

Estimating the effects of family structure is complicated by the
possibility of family structure being endogenous. The method-
ological problem is in determining whether differences in out-
comes between children living with different kinds of families are
causal because the adults establishing alternative household struc-
tures may differ in other regards. If families that become disrupted
or women who opt to establish single-mother households have a
particular characteristic that would have differentially inºuenced
childhood outcomes regardless of household structure, the attribu-
tion of different outcomes solely on the basis of household struc-
ture would be misleading. Evidence that children raised in dys-

16 Moehling, “Family Structure”; Margo, Race and Schooling. David Popenoe, Life without
Father: Compelling New Evidence that Fatherhood and Marriage are Indispensable for the Good of
Children and Society (New York, 1996), among others, notes that family disruptions due to di-
vorce or voluntary separation have broader and deeper negative effects on child outcomes
than disruptions due to death. The evidence concerning stepparents is mixed, but the weight
of it suggests that stepparents are not a full replacement for biological parents.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

46 | HOWARD BODENHORN

functional families were not much different than children raised in
disrupted families would provide support for the contention that
unobserved parental characteristics are more important
than
household structure per se in explaining child achievement. The
issue of dysfunction versus disruption deserves further attention,
but it is well beyond the scope of this study, which follows the lit-
erature in accepting that household structure per se has notable
consequences.17

If unobserved factors that lead to the establishment of a non-
traditional household are correlated with factors inºuencing the
choice to enroll a child in school or send him or her into the labor
market, failing to account for this endogeneity will lead to biased
parameter estimates; the appropriate method would be instrumen-
tal variable estimation. In the present context, however, it is un-
clear what such instruments might be. Serious concerns about the
effects of endogeneity may be allayed by the results of IV estima-
tion by Manski et al. and McLanahan and Sandefur, which ªnd
that treating household structure as exogenous does not lead to
inappropriate inferences. These precedents, however, do not pro-
vide a license to ignore the complications arising from potential
endogeneity in the present study. But if the potential biases are
small, and the resulting claims are not exorbitant, treating house-
hold structures as exogenous is methodologically acceptable.18

household structure and school attendance The effects
of living in a nontraditional household on child and youth school
attendance are estimated from probit regressions on the dichoto-
mous school-attendance variable, according to the following spec-
iªcation:

(School Attendance)jhkw

(cid:2) (cid:3) (cid:4) (cid:5) Sjh
where j indexes the child; h indexes the household; k indexes the
city; and w indexes the city ward. The vector S includes dichoto-
mous variables for single mothers, single fathers, and children liv-

(cid:4) (cid:6) Xjh

(cid:4) (cid:9)j,

(cid:7)

(cid:8)

hk

jw

17 David T. Ellwood and Christopher Jencks, “The Spread of Single-Parent Families in the
United States since 1960,” working paper (Cambridge, Mass., 2002); McLanahan and
Bumpass, “Intergenerational Consequences.”
18 Charles F. Manski et al., “Alternative Estimates of the Effect of Family Structure during
Adolescence on High School Graduation,” Journal of
the American Statistical Association,
LXXXVII (1992), 25–37; McLanahan and Sandefur, Growing Up.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 47

ing with neither parent (the omitted variable is children residing
with both parents). The vector X includes a number of additional
individual and family controls,
including child’s age, head of
household’s age, literacy and nativity; (cid:7) controls for the city of res-
idence; and (cid:8) controls for the census ward in which the household
resided. Separate regressions are estimated for three age cohorts
(six to eight years, nine to thirteen years, and fourteen to sixteen
years) and two races. Effects are estimated for all children (boys
and girls) together. As reported in Table 1, attendance rates by age
cohort and race were nearly identical for both sexes, and prelimi-
nary separate regressions for boys and girls did not yield sig-
niªcantly different coefªcients. The ªnal speciªcations include a
dummy variable for boys, but the estimated effect is virtually nil
and statistically insigniªcant for white children. It is small and only
sometimes signiªcant for black children.

It is important to estimate household-structure effects after
controlling for potential household resource effects. The observed
negative outcomes for children raised in nontraditional homes are
often thought the result of resource constraints faced by single-
mother households. Hence, understanding the consequences of
single parenthood on child outcomes demands separate controls
for resource effects. The baseline speciªcations include no re-
source variables and are reported for comparative purposes only.
All discussions of the consequences of household structure are
drawn only from speciªcations that include resource controls. In
addition to household structures, the baseline models include
child’s age and its square, a dummy equal to one if the child is
male, the number of residents in the household, the number of
siblings between six and sixteen years, and a dummy variable for
Baltimore residence. For each age cohort and race, a second set of
equations are estimated that add several household-resource mea-
sures, including literacy status of household head, his or her age
and its square, a dummy variable indicating home ownership, and
a measure of household head’s occupational status. An alternative
speciªcation replaces
the home-ownership and occupational-
status variables with total household wealth. A third speciªcation
adds a vector of 30 dichotomous variables to capture differential
access to schools across thirty city wards.

Given that tables 3 and 4 report household-structure results
from thirty-six separate speciªcations, the discussion focuses on

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

48 | HOWARD BODENHORN
the estimates provided by the full speciªcations (Base (cid:4) Resource
(cid:4) Ward controls). Tables 3 and 4 report the estimated marginal
changes (from 0 to 1) from probit regressions. Standard errors cor-
rected for non-independence (clustering) across households are re-
ported in parentheses. Resource controls in Table 3 are home
ownership and occupational status; in Table 4, the resource con-
trol is total household wealth.

The results reveal some intriguing patterns in school atten-
dance. Marginal effects reported in Table 3 for white children be-
tween six and eight years of age reveal that living with a single
mother or single father had a small and insigniªcant impact on
school attendance. For black children, however, living with a sin-
gle mother reduced attendance by a statistically signiªcant 7 per-
cent. Although the point estimate for the consequence of single
fatherhood is larger, it is not signiªcantly different from zero. For
the youngest school-age black children, living with neither parent
reduced school attendance by 9 percent, relative to young black
children residing in a two-parent household. The estimated mar-
ginal impacts reported in Table 4, which include alternative re-
source controls, are nearly identical to those in Table 3. Young
black children living with a single mother or living with neither
parent were about 7 to 9 percent less likely to attend school than
their counterparts living with two parents.

For children in the mid-nineteenth century’s prime school-
age years (nine to thirteen), residing in a household with a single
mother had negligible effects on school attendance, a result that
holds for both blacks and whites. Blacks were less likely to attend
school at any age, but single parenthood did not compound the
negative consequences of race for children in this age group. This
ªnding is consistent with Vinovskis’ observation concerning a
powerful social norm that children in the prime attendance years
should be sent to school. Although poverty and lesser access lim-
ited black school attendance to about half of the white rate for
nine- to thirteen-year-olds, the attendance rate of 40 percent rep-
resented is a testament to the broad recognition among free blacks
that long-term economic well-being was heavily dependent on
education.

Like Vinovskis’ study, this article documents the popular view
that schooling in the teenage years depended on wealth and social
standing. Attendance for children aged fourteen to sixteen falls to
one-third to one-half the attendance rate of nine- to thirteen-

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 49

year-olds. Moreover, single motherhood appears to have been a
powerful deterrent to continued attendance at older ages. White
youths residing in a single-mother household were about 8 to 9
percent less likely to attend school than youths in two-parent
households, and single-motherhood reduced school attendance of
black youths by about 3 to 4 percent, after the implementation of
resource and ward controls.

Murray’s study of Charleston reveals that parents, even indi-
gent mothers, demanded that their children receive a proper edu-
cation. Most did. Were the same pressures placed on families with
foster or orphaned children or young servants? The 1860 census
indicated whether a child resided with neither parent and some-
times whether a child was a live-in servant, but it offers no way to
assign “foster” status to children not acting as servants. Thus, all
children living with neither parent are grouped together, whether
they were servants, apprentices, orphaned, or abandoned children
in the care of family or friends. Given that a resident servant was
probably treated in a different fashion than, say, an orphaned
niece, this grouping may not capture the nuances of foster care at
mid-century. It is, therefore, important not to read too much into
the ªndings, but, given the large proportion of children living
with neither parent, it is also important to understand the upshot
of this household structure even if imperfectly.19

At every age, residing with neither parent was associated with
a much lower school-attendance rate. White children from six to
eight years old residing with neither parent attended school at rates
15 percent below those of children in two-parent households.
Black attendance rates were about 9 percent lower at these young
ages. In the prime school-age years, the attendance rate for whites
was 22 percent lower than for children in two-parent households;
for blacks it was 16 percent lower. Finally, white children aged
fourteen to sixteen who lived with neither parent attended at a
rate 31 percent below that of children in two-parent households.
For blacks, the neither-parent effect was less pronounced than for
whites at older ages, falling to about 6 percent.

Dramatically lower attendance rates among white teens living
with neither parent may not be indicative of sharply reduced hu-
man-capital formation among this group relative to children living

John Murray, “Family, Literacy, and Skill Training in the Antebellum South: Historical

19
Longitudinal Evidence from Charleston,” Journal of Economic History, LXIV (2004), 773–799.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

6
1
–
4
1

s
e
g
a

3
1
–
9

s
e
g
a

8
–
6

s
e
g
a

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

e
g
A

d
n
a

e
c
a
R
y
b

e
c
n
a
d
n
e
t
t

A

l

o
o
h
c
S

n
o

e
r
u
t
c
u
r
t
S

l

d
o
h
e
s
u
o
H

f
o

s
t
c
e
f
f

E

d
e
t
a
m

i
t
s
E

3

e
l
b
a
T

‡
)
7
4
0
.
(

6
0
.
0
(cid:10)

)
5
5
0
.
(

1
3
.
0
(cid:10)

*
)
1
3
0
.
(

8
0
.
0
(cid:10)

%
0
.
8
3

4
.
9
2

‡
)
6
4
0
.
(

5
0
.
0
(cid:10)

)
8
5
0
.
(

8
2
.
0
(cid:10)

*
)
7
3
0
.
(

8
0
.
0
(cid:10)

%
0
.
8
3

2
.
9
2

†
)
4
4
0
.
(

4
0
.
0
(cid:10)

)
5
5
0
.
(

0
3
.
0
(cid:10)

0
1
.
0
(cid:10)

)
2
3
0
.
(

5
0
.
0
(cid:10)

2
0
.
0
(cid:10)

)
3
4
0
.
(

2
2
.
0
(cid:10)

)
3
3
0
.
(

3
0
.
0
(cid:10)

4
0
.
0
(cid:10)

)
8
3
0
.
(

4
2
.
0
(cid:10)

*
)
3
3
0
.
(

*
)
9
5
0
.
(

*
)
4
5
0
.
(

%
0
.
8
3

6
.
3
3

%
5
.
7
7

9
.
8
6

%
5
.
7
7

4
.
2
7

‡
)
3
3
0
.
(

2
0
.
0
(cid:10)

)
8
3
0
.
(

5
2
.
0
(cid:10)

*
)
3
5
0
.
(

5
0
.
0
(cid:10)

%
5
.
7
7

6
.
4
7

)
7
4
0
.
(

0
0
.
0

3
0
.
0

)
7
8
0
.
(

)
0
5
0
.
(

4
1
.
0
(cid:10)

%
3
.
7
5

0
.
0
5

)
4
4
0
.
(

5
0
.
0

)
9
4
0
.
(

5
1
.
0
(cid:10)

2
0
.
0
(cid:10)

)
5
4
0
.
(

4
0
.
0

)
8
4
0
.
(

6
1
.
0
(cid:10)

3
0
.
0
(cid:10)

‡
)
8
7
0
.
(

†
)
8
7
0
.
(

t
n
e
r
a
p

r
e
h
t
i
e
N

r
e
h
t
a
f

e
l
g
n
i
S

r
e
h
t
o
m
e
l
g
n
i
S

n
e
r
d
l
i

h
c

e
t
i
h
W

%
3
.
7
5

9
.
2
5

%
3
.
7
5

6
.
7
5

e
c
n
a
d
n
e
t
t
a

l
a
u
t
c
A

k
c
a
l
b

h
t
i

W

s
c
i
t
s
i
r
e
t
c
a
r
a
h
c

)
9
1
0
.
(

1
0
.
0

)
6
3
0
.
(

6
0
.
0
(cid:10)

3
0
.
0
(cid:10)

)
1
2
0
.
(

3
0
.
0

)
8
3
0
.
(

6
0
.
0
(cid:10)

3
0
.
0
(cid:10)

†
)
0
2
0
.
(

2
0
.
0

)
7
3
0
.
(

7
0
.
0
(cid:10)

4
0
.
0
(cid:10)

)
6
3
0
.
(

0
0
.
0

)
5
5
0
.
(

6
1
.
0
(cid:10)

2
0
.
0
(cid:10)

)
4
3
0
.
(

1
0
.
0

)
6
5
0
.
(

8
1
.
0
(cid:10)

4
0
.
0
(cid:10)

†
)
9
1
0
.
(

†
)
2
2
0
.
(

†
)
1
2
0
.
(

*
)
9
3
0
.
(

*
)
8
3
0
.
(

%
0
.
2
1

3
.
9
1

%
0
.
2
1

0
.
6
1

%
0
.
2
1

8
.
1
1

%
1
.
7
3

6
.
9
4

%
1
.
7
3

4
.
8
3

)
3
3
0
.
(

2
0
.
0

)
7
5
0
.
(

9
1
.
0
(cid:10)

5
0
.
0
(cid:10)

*
)
7
3
0
.
(

%
1
.
7
3

4
.
8
3

†
)
6
3
0
.
(

9
0
.
0
(cid:10)

)
7
4
0
.
(

9
0
.
0
(cid:10)

†
)
9
3
0
.
(

7
0
.
0
(cid:10)

%
7
.
5
2

2
.
1
3

†
)
4
3
0
.
(

7
0
.
0
(cid:10)

)
0
5
0
.
(

2
1
.
0
(cid:10)

*
)
6
3
0
.
(

8
0
.
0
(cid:10)

%
7
.
5
2

7
.
7
2

†
)
4
3
0
.
(

7
0
.
0
(cid:10)

)
9
4
0
.
(

3
1
.
0
(cid:10)

*
)
5
3
0
.
(

8
0
.
0
(cid:10)

%
7
.
5
2

6
.
6
2

t
n
e
r
a
p

r
e
h
t
i
e
N

r
e
h
t
a
f

e
l
g
n
i
S

r
e
h
t
o
m
e
l
g
n
i
S

n
e
r
d
l
i

h
c

k
c
a
l
B

e
c
n
a
d
n
e
t
t
a

l
a
u
t
c
A

e
t
i
h
w
h
t
i

W

s
c
i
t
s
i
r
e
t
c
a
r
a
h
c

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

e
l
g
n
i
s

f
o

n
o
i
t
i
n
ª
e
d

r
o
f

x
i
d
n
e
p
p
A
e
e
S

.
)
s
n
o
i
s
s
e
r
g
e
r

t
i
b
o
r
p
m
o
r
f

s
t
c
e
f
f
e

l
a
n
i
g
r
a
m

(

s
u
t
a
t
s

l
a
n
o
i
t
a
p
u
c
c
o

d
n
a

p
i
h
s
r
e
n
w
o

e
m
o
h

e
d
u
l
c
n

i

s
l
o
r
t
n
o
C
e
c
r
u
o
s
e
R

.
1
0
.
0
(cid:11)
e
u
l
a
v
–
p

.
5
0
.
0
(cid:11)
e
u
l
a
v
–
p

.
0
1
.
0
(cid:11)
e
u
l
a
v
–
p

*

†

‡

s
e
t
o
n

l

,
e
r
u
t
c
u
r
t
s
d
o
h
e
s
u
o
h
o
t
n
o
i
t
i
d
d
a
n
I

.
s
e
l
b
a
i
r
a
v
y
m
m
u
d
r
o
f
1
o
t
0
m
o
r
f

e
g
n
a
h
c

f
o
s
t
c
e
f
f
e

l
a
n
i
g
r
a
m
d
e
t
a
m

i
t
s
e

s
t
r
o
p
e
r

e
l
b
a
T

.
t
n
e
r
a
p
r
e
h
t
i
e
n
d
n
a

,
r
e
h
t
a
f

e
l
g
n
i
s

,
r
e
h
t
o
m

y
m
m
u
d

a

d
n
a

,

n
e
e
t
x
i
s

d
n
a

x
i
s

s
e
g
a

n
e
e
w
t
e
b

s
g
n

i
l
b
i
s

f
o

r
e
b
m
u
n

e
h
t

l

,
d
o
h
e
s
u
o
h

e
h
t

n

i

s
t
n
e
d
i
s
e
r

f
o

r
e
b
m
u
n

e
h
t

,
e
r
a
u
q
s

s
t
i

d
n
a

e
g
a

s
’
d
l
i

h
c

e
d
u
l
c
n

i

s
n
o
i
s
s
e
r
g
e
r

e
s
a
b

l

d
a
e
h
d
o
h
e
s
u
o
h
e
h
t

f
i

e
n
o
o
t

l
a
u
q
e

e
l
b
a
i
r
a
v

y
m
m
u
d
a

s
n
o
i
s
s
e
r
g
e
r

e
s
a
b
e
h
t
o
t
d
e
d
d
a

s
n
o
i
s
s
e
r
g
e
r

e
c
r
u
o
s
e
r

e
h
T

.
e
r
o
m

i
t
l
a
B
n

i
d
e
v
i
l

d
l
i

h
c

e
h
t

f
i

e
n
o
o
t

l
a
u
q
e

e
l
b
a
i
r
a
v

e
l
b
a
i
r
a
v
y
m
m
u
d
a

l

,
d
o
h
e
s
u
o
h
e
h
t

f
o
d
a
e
h
e
h
t

f
o
e
r
a
u
q
s

s
t
i
d
n
a

e
g
a

e
h
t

,
e
t
a
r
e
t
i
l

l

s
a
w
d
o
h
e
s
u
o
h
e
h
t

f
o
d
a
e
h
e
h
t

f
i

e
n
o
o
t

l
a
u
q
e

e
l
b
a
i
r
a
v
y
m
m
u
d
a

,
t
n
a
r
g
i
m
m

i

n
a

s
a
w

l
a
n
o
i
t
a
p
u
c
c
o

n
a

d
n
a

,
)
h
t
l
a
e
w
e
t
a
t
s
e

l
a
e
r

r
o
f

e
u
l
a
v

e
v
i
t
i
s
o
p

a

d
e
t
r
o
p
e
r

l

d
o
h
e
s
u
o
h

e
h
t

f
i

e
n
o

s
l
a
u
q
e

t
i

,
s
i

t
a
h
t
(

e
m
o
h

n
w
o

r
i
e
h
t

d
e
n
w
o

y
l
i

m

a
f

e
h
t

f
i

e
n
o

o
t

l
a
u
q
e

d
n
a

e
r
o
m

i
t
l
a
B
n

i

s
d
r
a
w

s
u
s
n
e
c

e
n
o
–
y
t
r
i
h
t

e
h
t

f
o
y
t
r
i
h
t

r
o
f

s
e
l
b
a
i
r
a
v
y
m
m
u
d
d
d
a

s
n
o
i
s
s
e
r
g
e
r
d
r
a
w
e
h
T

.
)
n
o
i
t
c
u
r
t
s
n
o
c

s
t
i

r
o
f
x
i
d
n
e
p
p
A
e
h
t

e
e
s
(

r
o
e
l
b
a
i
r
a
v
s
u
t
a
t
s

i
e
s

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

.
s
e
s
e
h
t
n
e
r
a
p

n

i

l

s
d
o
h
e
s
u
o
h

f
o

e
c
n
e
d
n
e
p
e
d
n
i
–
n
o
n

r
o
f

d
e
t
c
e
r
r
o
c

s
r
o
r
r
e

d
r
a
d
n
a
t
s

t
s
u
b
o
R

.
s
n
a
e
l
r

O
w
e
N

6
1
–
4
1

s
e
g
a

3
1
–
9

s
e
g
a

8
–
6

s
e
g
a

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

s
d
r
a
w

s
e
c
r
u
o
s
e
r

e
s
a
b

e
g
A

d
n
a

e
c
a
R
y
b

e
c
n
a
d
n
e
t
t

A

l

o
o
h
c
S

n
o

e
r
u
t
c
u
r
t
S

l

d
o
h
e
s
u
o
H

f
o

s
t
c
e
f
f

E

d
e
t
a
m

i
t
s
E

4

e
l
b
a
T

‡
)
7
4
0
.
(

6
0
.
0
(cid:10)

)
5
5
0
.
(

1
3
.
0
(cid:10)

*
)
0
3
0
.
(

9
0
.
0
(cid:10)

%
0
.
8
3

8
.
7
2

‡
)
9
1
0
.
(

1
0
.
0

†
)
9
1
0
.
(

)
4
3
0
.
(

6
0
.
0
(cid:10)

4
0
.
0
(cid:10)

%
0
.
2
1

8
.
1
2

†
)
5
4
0
.
(

5
0
.
0
(cid:10)

)
8
5
0
.
(

9
2
.
0
(cid:10)

*
)
7
3
0
.
(

9
0
.
0
(cid:10)

%
0
.
8
3

9
.
8
2

‡
)
0
2
0
.
(

1
0
.
0

†
)
1
2
0
.
(

)
5
3
0
.
(

6
0
.
0
(cid:10)

4
0
.
0
(cid:10)

%
0
.
2
1

9
.
0
2

†
)
5
4
0
.
(

4
0
.
0
(cid:10)

)
5
5
0
.
(

0
3
.
0
(cid:10)

*
)
3
3
0
.
(

0
1
.
0
(cid:10)

%
0
.
8
3

6
.
3
3

)
2
3
0
.
(

4
0
.
0
(cid:10)

3
0
.
0
(cid:10)

)
3
4
0
.
(

2
2
.
0
(cid:10)

*
)
9
5
0
.
(

%
5
.
7
7

4
.
9
6

‡
)
3
3
0
.
(

2
0
.
0
(cid:10)

)
8
3
0
.
(

5
2
.
0
(cid:10)

*
)
4
5
0
.
(

5
0
.
0
(cid:10)

%
5
.
7
7

1
.
2
7

‡
)
4
3
0
.
(

2
0
.
0
(cid:10)

)
8
3
0
.
(

5
2
.
0
(cid:10)

*
)
4
5
0
.
(

5
0
.
0
(cid:10)

%
5
.
7
7

6
.
4
7

)
6
4
0
.
(

3
0
.
0

0
0
.
0

‡
)
9
8
0
.
(

)
0
5
0
.
(

5
1
.
0
(cid:10)

%
3
.
7
5

9
.
4
5

†
)
0
2
0
.
(

2
0
.
0

)
7
3
0
.
(

7
0
.
0
(cid:10)

4
0
.
0
(cid:10)

)
6
3
0
.
(

0
0
.
0

)
5
5
0
.
(

6
1
.
0
(cid:10)

3
0
.
0
(cid:10)

)
4
3
0
.
(

1
0
.
0

)
6
5
0
.
(

8
1
.
0
(cid:10)

5
0
.
0
(cid:10)

†
)
1
2
0
.
(

*
)
9
3
0
.
(

*
)
8
3
0
.
(

%
0
.
2
1

8
.
1
1

%
1
.
7
3

1
.
0
5

%
1
.
7
3

7
.
3
4

)
3
3
0
.
(

2
0
.
0

)
7
5
0
.
(

9
1
.
0
(cid:10)

5
0
.
0
(cid:10)

*
)
7
3
0
.
(

%
1
.
7
3

4
.
8
3

†
)
5
3
0
.
(

9
0
.
0
(cid:10)

)
7
4
0
.
(

9
0
.
0
(cid:10)

†
)
9
3
0
.
(

8
0
.
0
(cid:10)

%
7
.
5
2

1
.
2
3

)
5
4
0
.
(

5
0
.
0

)
9
4
0
.
(

6
1
.
0
(cid:10)

3
0
.
0
(cid:10)

†
)
9
7
0
.
(

%
3
.
7
5

2
.
7
5

*
)
3
3
0
.
(

7
0
.
0
(cid:10)

)
0
5
0
.
(

3
1
.
0
(cid:10)

*
)
6
3
0
.
(

9
0
.
0
(cid:10)

%
7
.
5
2

1
.
9
2

)
5
4
0
.
(

4
0
.
0

)
8
4
0
.
(

6
1
.
0
(cid:10)

3
0
.
0
(cid:10)

†
)
8
7
0
.
(

%
3
.
7
5

6
.
7
5

†
)
4
3
0
.
(

7
0
.
0
(cid:10)

)
9
4
0
.
(

3
1
.
0
(cid:10)

*
)
5
3
0
.
(

8
0
.
0
(cid:10)

%
7
.
5
2

6
.
6
2

t
n
e
r
a
p

r
e
h
t
i
e
N

r
e
h
t
a
f

e
l
g
n
i
S

r
e
h
t
o
m
e
l
g
n
i
S

n
e
r
d
l
i

h
c

e
t
i
h
W

e
c
n
a
d
n
e
t
t
a

l
a
u
t
c
A

k
c
a
l
b

h
t
i

W

s
c
i
t
s
i
r
e
t
c
a
r
a
h
c

r
e
h
t
o
m
e
l
g
n
i
S

n
e
r
d
l
i

h
c

k
c
a
l
B

t
n
e
r
a
p

r
e
h
t
i
e
N

r
e
h
t
a
f

e
l
g
n
i
S

e
c
n
a
d
n
e
t
t
a

l
a
u
t
c
A

e
t
i
h
w
h
t
i

W

s
c
i
t
s
i
r
e
t
c
a
r
a
h
c

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

d
n
a

,
r
e
h
t
a
f

e
l
g
n
i
s

,
r
e
h
t
o
m
e
l
g
n
i
s

f
o

n
o
i
t
i
n
ª
e
d

r
o
f

x
i
d
n
e
p
p
A
e
e
S

.
)
s
n
o
i
s
s
e
r
g
e
r

t
i
b
o
r
p
m
o
r
f

s
t
c
e
f
f
e

l
a
n
i
g
r
a
m

(

h
t
l
a
e
w
d
o
h
e
s
u
o
h

l

l
a
t
o
t

s
i

l

o
r
t
n
o
c

e
c
r
u
o
s
e
R

s
e
t
o
n

e
d
u
l
c
n

i

s
n
o
i
s
s
e
r
g
e
r

e
s
a
b

,
e
r
u
t
c
u
r
t
s

l

d
o
h
e
s
u
o
h

o
t

n
o
i
t
i
d
d
a

n
I

.
s
e
l
b
a
i
r
a
v

y
m
m
u
d

r
o
f

1

o
t

0
m
o
r
f

e
g
n
a
h
c

f
o

s
t
c
e
f
f
e

l
a
n
i
g
r
a
m
d
e
t
a
m

i
t
s
e

s
t
r
o
p
e
r

e
l
b
a
T

.
t
n
e
r
a
p

r
e
h
t
i
e
n

f
i

a

e
n
o

o
t

l
a
u
q
e

e
l
b
a
i
r
a
v

y
m
m
u
d

a

d
n
a

,

n
e
e
t
x
i
s

d
n
a

x
i
s

s
e
g
a

n
e
e
w
t
e
b

s
g
n

i
l
b
i
s

f
o

r
e
b
m
u
n

e
h
t

l

,
d
o
h
e
s
u
o
h

e
h
t

n

i

s
t
n
e
d
i
s
e
r

f
o

r
e
b
m
u
n

e
h
t

,
e
r
a
u
q
s

s
t
i

d
n
a

e
g
a

s
’
d
l
i

h
c

,
t
n
a
r
g
i
m
m

i

n
a

s
a
w
d
a
e
h

l

d
o
h
e
s
u
o
h

e
h
t

f
i

e
n
o

o
t

l
a
u
q
e

e
l
b
a
i
r
a
v

y
m
m
u
d

a

s
n
o
i
s
s
e
r
g
e
r

e
s
a
b

e
h
t

o
t

d
e
d
d
a

s
n
o
i
s
s
e
r
g
e
r

e
c
r
u
o
s
e
r

e
h
T

.
e
r
o
m

i
t
l
a
B
n

i

d
e
v
i
l

d
l
i

h
c

e
h
t

e
h
T

.

h
t
l
a
e
w
d
o
h
e
s
u
o
h

l

l
a
t
o
t

e
h
t

d
n
a

l

,
d
o
h
e
s
u
o
h
e
h
t

f
o
d
a
e
h
e
h
t

f
o
e
r
a
u
q
s

s
t
i

d
n
a

e
g
a

e
h
t

,
e
t
a
r
e
t
i
l

l

s
a
w
d
o
h
e
s
u
o
h
e
h
t

f
o
d
a
e
h
e
h
t

f
i

e
n
o
o
t

l
a
u
q
e

e
l
b
a
i
r
a
v

y
m
m
u
d

–
e
d
n
i
–
n
o
n
r
o
f

d
e
t
c
e
r
r
o
c

s
r
o
r
r
e

d
r
a
d
n
a
t
s

t
s
u
b
o
R

.
s
n
a
e
l
r

O
w
e
N
d
n
a

e
r
o
m

i
t
l
a
B
n

i

s
d
r
a
w

s
u
s
n
e
c

e
n
o
–
y
t
r
i
h
t

e
h
t

f
o
y
t
r
i
h
t

r
o
f

s
e
l
b
a
i
r
a
v

y
m
m
u
d

d
d
a

s
n
o
i
s
s
e
r
g
e
r

d
r
a

W

.
s
e
s
e
h
t
n
e
r
a
p

n

i

l

s
d
o
h
e
s
u
o
h

f
o

e
c
n
e
d
n
e
p

.
1
0
.
0
(cid:11)
e
u
l
a
v
–
p

.
5
0
.
0
(cid:11)
e
u
l
a
v
–
p

.
0
1
.
0
(cid:11)
e
u
l
a
v
–
p

s
e
i
l
p
m

i

s
e
i
l
p
m

i

s
e
i
l
p
m

i

*

†

‡

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

54 | HOWARD BODENHORN

with two parents. Records from Baltimore reveal that children ap-
prenticed by their fathers were less likely to demand education or
schooling covenants in apprenticeship indentures. Pauper appren-
tices, however, were more likely to have such clauses included in
their indenture contracts, suggesting that additional human-capital
formation during the teenage years took the form of skill training
after the completion of primary education. The interpretation is
less sanguine for black children, who were apprenticed into less
lucrative occupations and were less likely to have completed a pri-
mary education at the time of the apprenticeship. Furthermore,
Maryland law relieved masters from the responsibility of sending
black apprentices to school. Apprenticeship alone does not explain
low rates of school attendance among teenagers; nor does it ex-
clude alternative explanations of relatively low attendance rates
among youths living with neither parent. Moehling’s study sug-
gests that non-kin foster children in 1910 attended at rates about
13 to 20 percent less than children from two-parent families; re-
lated foster children attended at rates about 6 to 15 percent less.
The results in this article are broadly consistent with her ªndings,
namely, that the long-term costs, in terms of foregone human-
capital formation, to children not residing with two parents or,
absent that, kin of some type were quite high.20

The pattern of results reported in tables 3 and 4 supports the
conclusion that single motherhood reduced school attendance
among young black children (six to eight years old) by about 7 to
9 percent, relative to children in two-parent households. Being
black and living with a single mother also reduced school atten-
dance of children aged fourteen to sixteen, though the effect of
single motherhood is less pronounced for them than for younger
children. That single motherhood had negligible effects on school
attendance during the prime school age of nine to thirteen attests
to the powerful social norm of nearly universal primary education
for whites and a less powerful, but still strong, impulse toward ed-
ucating black children. Nevertheless, single motherhood in black
households was associated with less lifetime education because
black children living with single mothers apparently started school
later and left school earlier than children living with two parents.

20 Bodenhorn, “Just and Reasonable Treatment: Racial Differences in the Terms of Pauper
Apprenticeship in Antebellum Maryland,” National Bureau of Economic Research working
paper 9752 (2003); Moehling, “Family Structure.”

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 55

For white children, single motherhood was associated with leav-
ing school early but not with reduced attendance at early ages.

White teens residing in mother-only households were proba-
bly expected to enter the labor force to provide income for the
household (an issue discussed below). The less pronounced effect
of single motherhood on black teens may be due to two effects.
First, since black teens were much less likely than white teens
to attend school regardless of household structure, the result of
single motherhood was less pronounced—though a reduction in
attendance of 4 percent when the mean attendance rate was only
12 percent seems substantial. Second, given the low wages paid to
black workers, the opportunity cost of remaining in school during
the mid-teen years may have been lower for blacks than whites.
Nonetheless, it seems likely that poor single mothers would have
sent teens into the labor market more often to contribute toward
the maintenance of the household. The next section investigates
the extent of that behavior.21

the effects of household structure on early labor-force
participation As reported in Table 1, only a small fraction of
females between the ages of fourteen and sixteen participated
in the labor force. Hence, the focus in this section is on males,
about 25 percent of whom had sufªcient participation in the labor
force to merit acknowledgment by census enumerators. The ef-
fects, estimated from probit regressions on the dichotomous labor-
force participation variable, are estimated from the following
speciªcation:

(cid:2) (cid:3) (cid:4) (cid:5) Sjh

(cid:4) (cid:9)j,
(Labor Force Participation)jhkw
where j indexes the child; h indexes the household; k indexes the
city; and w indexes the city ward. The independent variables are
the same as those used in estimating the effects of household struc-
ture on school attendance. Three separate speciªcations are esti-
mated: a baseline model without resource effects, a second model

(cid:4) (cid:6) Xjh

(cid:7)

(cid:8)

hk

jw

21 Bodenhorn, “Just and Reasonable,” reports that local ofªcers of the court frequently
placed children living in mother-only households in apprenticeships when the mother was
considered incapable of transforming an adolescent into a productive adult. But some appren-
ticeships were apprenticeships in name only, like those of children put to work in proto-
factories or as servants who continued to live at home. It is not clear how much income such
apprenticed children would provide to the household, though some received modest cash
wages.

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

56 | HOWARD BODENHORN

that includes measures of household resources, and a full spec-
iªcation that also includes dummy variables for wards. Finally,
separate effects are estimated for black and white youth.22

Table 5 reports the results of twelve probit-regression spec-
iªcations. The full speciªcations for white male youths reveal that
living with a single mother increased labor-force participation by
11 percent, a result that is robust to alternative measures of house-
hold resources. Moreover, the single-mother effect increases in
magnitude and statistical signiªcance when controls for household
resources and wards are included. For white male youth, living
with a single father or with neither parent had no statistically
signiªcant effect on labor-force participation. Single motherhood
was associated with a higher incidence of early labor-force attach-
ment among white male youth.

The association between household structure and labor-force
attachment among blacks differs sharply from the association for
whites. Single motherhood had a small and statistically insigniªcant
effect on labor-force participation among fourteen-to sixteen-
year-old black youths. Similarly, single fatherhood had no discern-
ible effect on labor-force attachment. Living with neither parent,
however, had a meaningful and signiªcant effect on the employ-
ment of black male youths. Those living with neither parent were
about 16 to 17 percent more likely than black youths living with
both parents to have regular attachment to the labor force.

Although the consequences of

family structure differ for
blacks and whites, the near equality of labor-force participation
rates across races is notable. As is clear from lines 4 and 5 in
Table 5, eliminating the racial differences in youth characteristics
leads to small changes in predicted labor-force participation. If
white youths had the characteristics of black youths, for example,
their labor-force participation rate would have increased from 23.2
to 25.5 percent when conditioned on total household wealth.
Giving black youths white characteristics would have increased
black participation from 28.1 to 32.2 percent, conditioning on
household wealth. Conditioning on home ownership and occupa-

22
It is difªcult to reconcile the low female participation rate observed in this sample with
the results reported in Claudia Goldin, Understanding the Gender Gap: An Economic History of
American Women (New York, 1990). The discrepancy probably stems from this study’s focus
on girls aged fourteen to sixteen, whereas Goldin considers single women from ªfteen to
twenty-four years of age. Goldin ªnds considerably higher labor-market participation rates
but ironically concludes that girls in the market labor force achieved greater autonomy and
more leisure time than girls laboring in the home (50–55).

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

SINGLE PARENTHOOD AND CH ILD H OOD O UTCOM ES | 57

tional status of the household head increases participation rates for
both races about 5 to 7 percentage points. These ªndings are con-
sistent with those of Moehling, who found relatively modest racial
effects for male youth employment in 1910 after including a full
panel of controls.23

Interestingly, the labor-force participation regressions show
that racial differences in participation were driven more by family
structure than by race itself. Although single motherhood had a
large effect on white teen participation, it had no meaningful ef-
fect on black youth participation. Living with neither parent,
however, had a meaningful effect on black youth employment but
no effect on white employment. These results suggest two conclu-
sions: (1) Unlike orphaned or abandoned black children, orphaned
or abandoned white children apparently had recourse to some
kind of foster care that protected them from work at an early age;
(2) the powerful social norm of sending nine- to thirteen-year-
olds to school was apparently supplemented in the urban South by
one against sending young teens into the labor force. About 40
percent of white fourteen- to sixteen-year-olds attended school,
and another 23 percent had regular labor-force attachment, leav-
ing more than 33 percent of them regularly unoccupied. About 60
percent of black youth were apparently unoccupied on a regular
basis. A question for future research is accounting for these seem-
ingly unoccupied youths. Some of them undoubtedly performed
household chores. What else did they do with their time?

Evidence from Baltimore’s jailhouse records supports the
old adage about idle hands and the devil’s workshop. Perhaps only
by coincidence were black youth twice as likely to be idle and to
be arrested and jailed in Baltimore from 1854 to 1861. But this
issue is worthy of further study. If the connection is indeed
meaningful, the social costs of idleness were considerably higher
than they already were by denying black youth equal educational
access.24

Family structure mattered in the nineteenth-century urban South
just as much as it does today. It was an important determinant

23 The test z-statistic for testing the equality of the means is 1.75 (p value, 0.08). Thus, we
fail to reject the null hypothesis of equal means at the standard 1% or 5% levels. Moehling,
“Family Structure.”
24
half the arrest rate of black youths (13.4 and 10.5 per 1,000).

In 1858 and 1859, the white arrest rate (5.1 and 4.5 per 1,000 youths) was less than one-

l

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

p

:
/
/

d
i
r
e
c
t
.

m

i
t
.

e
d
u

/
j
i

/

n
h
a
r
t
i
c
e
–
p
d

l

f
/

/

/

/

3
8
1
3
3
1
6
9
8
0
0
2

/
j
i

.

.

n
h
2
0
0
7
3
8
1
3
3
p
d

.

.

.

f

b
y
g
u
e
s
t

t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

s
d
r
a
w