in
3:\.v\o
1 E x. 7 13
1:}: ,c;

:11.

6.3 .
Vlaitiz
7K it Ob.

"g 32A
3.1.12.3}... .
9 1:3 1:...
.5: 2
1..

nth-Vi... 1.I3JG?..\¢I;.

It"! 11!; #397}!
.n

)7‘.

:9 ‘

.r. til I
1.221.31‘ s.
sifterrail.
as t. ..

15 .
.lv. .21....
v.11} v

7.3.6): .151!

1 172.2%}: .5:

ivaavnllav.

c :- . o i.

1:... I ,

3.1.5:...

.31.}, 3...?!
13:83.???31IE.
5:23....)3’31 3X J3....£ .5

:Dr til I. . :r
1%.. :3 it.) 4...; .1...-
s 19..-...
2...}. .34! ‘
:t}:).§..lt
t t..)£...c.|:r:.
a? .

x13: , .1...
311,311.: 0-
1 $1.2..1 x.
.31 $53.32:

a l: .3 .
{is : 3.92.
(”vial » 1.". Pool. {)ol .5 e
.3. a: .. . .. a}. . I11:
1. 93’} !lox.i34}vt.h.q.a\cy01 JEN-1......ial a)... .
2.7 I). s..x.4.|ala7.4~1915.... as: I v31..-:p... .
61.2.. Eaasnntzs. (:5): 2.?! ,r .
:1. . 1 .

13.3.1110?! O

3.11.1! a!:5.lllf.bt»ll.!

is! {'3' 02¢ s. l-
, nI‘btl ‘5).(1

. r! 7: I)..|.|f.‘,
[JO-3.51.»
. 5..

 

.CI l\n6

 

Q 5 7 o 3 7 9 y
Illlll,.llg!3 mysgww’m L,
LIBRARY
Misc-n State
University

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

Tenancy and the Domestic Domain: Fertility
and Household Organization among Postbellum
Mississippi Tenant Farmers

presented by

Kim Arbogast McBride

has been accepted towards fulfillment
of the requirements for

Ph. D. degreein Anthropology

 

 

Major professor

 

mews], 17 70

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

D

 

 

ATE DUE DATE DUE DATE DUE

#1

 

 

 

 

 

 

 

 

 

 

Q.
"B!
3
Q9
C)
N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I: An Mrmdive ActiorVEquel Opportunity lnditution

 

 

 

TENANCY AND THE DOMESTIC DOMAIN: FERTILITY AND HOUSEHOLD
ORGANIZATION AMONG POSTBELLUM MISSISSIPPI TENANT FARMERS

By
Kim Arbogast McBride

A DISSSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Anthropology

1990

ABSTRACT

TENANCY AND THE DOMESTIC DOMAIN: FERTILITY AND HOUSEHOLD
ORGANIZATION AMONG POSTBELLUM MISSISSIPPI TENANT FARMERS

By
Kim Arbogast McBride

This dissertation uses household-level data from a late nineteenth to early
twentieth century rural community in northeastern Mississippi to examine relationships
among household structure, fertility, and tenant fanning. Theories of proto-
industrialization and semi-proletarianization are used as analogies to postbellum
Southern tenancy, which resembled proto-industrial societies in its attenuated control of
basic resources; intensification of labor due to debts and the landlord’s control of
farming; and household-based organization of labor. Concepts such as the domestic
mode of production are applied to relationships between Southern tenant households
and the larger economy. These concepts are then used to examine intensification of

both production and reproduction and their role in capital accumulation.

The data consists of a variety of documentary records. The main types of
documents used are land deeds, chattel deeds, estate records, population and agricultural
schedules from the federal census, and a private account ledger. The study was
accomplished by a variety of methods. General historical research and reconstruction of
landholding patterns and tenant-landlord relationships was used to provide a historical
narrative of the study area. A series of demographic and agricultural data bases were
constructed, and linked by each household. Analysis of the household focused on
household composition, life cycle stage, and the position of persons within households.
Analysis of marriage patterns and fertility was based upon calculation of a series of

measures, including child-women ratios, children born per year of marriage, age-specific

fertility rates, indices of the level and character of fertility (m and M), and singulate mean
age at marriage. The fertility analysis was supplemented by a sample of households from
the nearby county seat town, in 1910. Analysis of the agricultural system was
accomplished by summary statistics on wealth and crops.

Major findings include the differences in agricultural production according to tenure
and race; a general similarity between black and white household structure but increasing
nucleation over time; earlier age at marriage and of leaving the parental household for tenant
farmers; and variations in the level of fertility and family limitation according to racial,

tenure, and residential factors.

Capyright by
KIM ARBOGAST MCBRIDE
1990

ACKNOWLEDGMENTS

As with most large studies, this dissertation was completed with the help of
many persons. Part of the data was collected during Michigan State University
Museum's Tombigbee Historic Townsites Project, funded and administered by the U.
S. Army Corps of Engineers and the National Park Service. Tef Rodeffer of the
National Park Service provided administrative assistance and personal encouragement.
A good portion of the background historical data was ably collected by Winston W.
Way, Donna Windham, Cindy Parrish, and Randy Sparks, who also provided many
good times along the way. Thanks to Jim McClurken and Peggy Uland Anderson for
sharing the results of oral historical research on Vinton, and helping make the historical
work come alive. I thank many repositories in Mississippi, especially the Clay,
Lowndes, and Monroe County courthouses, Evans Memorial Library, the West Point
Library, Mississippi Department of Archives and History, and Special Collections,
Mississippi State University, for saving and curating the documentary sources, as well
as for their hospitality. Local historian Jack Elliottt, Jr. generously shared the results of
years of research in the study area. He and other local scholars like Rufus Ward
provided assistance only interested insiders could.

Family and friends contributed incredible patience over the long haul of analysis
and writing. Fellow students at Michigan State University were also important
supporters. Colleagues at the University of Kentucky have also provided needed
encouragement, as well as generous patience when dissertation obsessions lured me
away from duties there. I especially thank family members Charles and Nora Arbogast
and Beverley McBride for additional moral and physical support, including emergency
funds for a final data collection and loan of a personal computer up to the heaviest
number crunching. My husband Stephen McBride provided encouragement even while
going through the same trials himself.

 

Many professors at Michigan State University contributed to this research.
Much of my inspiration has been sustained by educational and enjoyable collaboration
on population issues with committee members Harry M. Raulet , Ann V. Millard, and
Anne C. Meyering. They provided an intellectual excitement and example that will be
hard to match elsewhere. Some of this work was supported by the Michigan State
University Center for the Advanced Study of International Development, as well as the
Department of Anthropology. Although not on my committee, Professor Nan E.
Johnson gave me a very solid introduction to demographic techniques and literature, as
well as welcome encouragement. From my very first meeting with him, Charles
Cleland, chair of my doctoral committee and also principal investigator of the
Tombigbee Historic Townsites Project, encouraged me to think of the long-term
research value of the data, and to apply it to broader issues. I especially thank him for
this example, as well as his patience and assistance in completing the research. This
assistance included editing drafts that were admittedly rough, navigating the burdens of
missed deadlines, and reminding me that the research was worth finishing.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES .
LIST OF ABBREVIATIONS

CHAPTER 1. INTRODUCTION
Introduction .

Questions of Theory .

Households within Southern tenancy.
Southern Historical Demography
Data and Methods .

Arrangement of Chapters

CHAPTER 2. THE CONTEXT OF SOUTHERN TENANCY
Introduction to Southern tenancy .
The Vinton Community

CHAPTER 3. HOUSEHOLD COMPOSITION AND LIFE CYCLE
Introduction

Household Composition

The Household Life Cycle

Positions within Households. .

Differences in Household Positions by Tenure

CHAPTER 4. MARRIAGE AND FERTILITY
Introduction .

Marital Status 1n Vinton

Serial Marriage and Complex Households

Age at First Marriage .

Analysis of Fertility: Introduction .

Own Children Estimates of Total Fertility Rates
Other Fertility Measures .

The Age-Specific Pattern of Fertility.

Illegitimacy

CHAPTER 5. AGRICULTURAL PRODUCTION, TENANCY, AND THE
HOUSEHOLD

Introduction .

Agricultural Production by Race and Tenure .

Agricultural Production by Household Type and Life Cycle . .

Other Relationships Between Production and Household Composition

vi

168

CHAPTER 6. CONCLUSIONS
Marriage and the Household Unit
Fertility

Final Thoughts on the Domestic Mode of Production.

APPENDIX. Categories and Classifications .
LIST OF REFERENCES

vii

183

LIST OF TABLES

Table 3.1. Miscellaneous household measures, 1880, expressed as means .
Table 3.2. Miscellaneous household measures, expressed as means, 1910 .
Table 3. 3. Household types, by race and tenure groups, 1880
Table 3. 4. Household types, be race and tenure groups,1910 .
Table 3. 5. Female-headed households, Vinton, 1880 and 1910, compared”
Table 3. 6. Demographic measures by household types and race/tenure groups,
1880, means. .
Table 3. 7. Household extension, by race and tenure groups, 1880 .
Table 3. 8. Household extension, by race and tenure groups, 1910 .
Table 3.9. Household expansion, by tenure and race groups, 1880 .
Table 3.10 Household expansion, by race and tenure groups, 1910.
Table 3.11. Household augmentation, by race and tenure groups, 1880
Table 3.12. Household augmentation, by race and tenure groups, 1910 .
Table 3.13. Demographic measures by stages and race/tenure group, 1880,
expressed as means .
Table 3.14. Household composition, 1880 .
Table 3.15. Relationship of individuals to head of household, Vinton,1880
Table 3.16. Position/relationship of individuals to head of household,
Vinton, 1910. .
Table 3.17. Household position, Vinton, 1910, by tenure
Table 4. 1. Marital Status, 1880 and 1910 compared. .
Table 4. 2. Percentage of age and tenure groups in various marital statuses,
Vinton,1910. .
Table 4. 3. Age at First Marriage, Vinton, 191.0 .
Table 4. 4. Status of children in households, used 1n 1910 fertility rates
Table 4. 5. Adjustments to children, own-children fertility rates, 1910
Table 4. 6. Adjustments to women, own-children fertility rates, 1910
Table 4. 7. Fertility measures, 1880 and 1910 . .
Table 4. 8. Numbers of children born by duration of marriage and
Other measures .
Table 4. 9. Comparison of various fertility measures, 1910
Table 4.10. Children born and surviving to women age 51+.
Table 4.11. Values of n(a) and v(a) for calculation of m and M
Table 4.12. Women with illegitimate births .
Table 5.1. Agricultural measures, 1880, expressed as means .
Table 5.2. Agricultural measures, crop mix and food-related,1880
expressed as means

120

123

129
131
132
139
145

150

Table 5.3. Agricultural measures by household types, 1880, expressed as means 153
Table 5. 4. Agricultural measures by household stages, 1880, expressed as means 155

Table 5.5. Percentage that own farm, by household life cycle stage .
Table 5.6. Tenure by age of household head, 1880 and 1910

Table 5.7. Correlation coefficients, agricultural analysis, 1880 .
Table 5.8. Summary of expenses and credits from Watson Ledger (1911)
Table 5. 9. Schedule of purchases from Vinton store, 1911- 1912

Table 5.10. Correlation coefficients, Watson ledger (1911) .

viii

166
166
170
176
179
179

Table A. 1. Household stages. . . . . . 208
Table A. 2. Consumer/worker equivalents . . . . . 210

ix

LIST OF FIGURES

Figure 2.1. Location of study area within black belt .

Figure 2.2. Location of study area and nearby communities .

Figure 2.3. Land acquisitions of W. E. Trotter, late 19th century
Figure 2.4. Number of Vinton store customers by location of farm .
Figure 2. 5. Number of adult black males by location of farm, 1870.
Figure 2. 6. New landowners 1n study area .

Figure 2. 7. Distribution of personal property, District 1, Clay County
Figure 3.1. Population structure, Vinton, black sample

Figure 3. 2. Population structure, Vinton, white sample .

Figure 3. 3. Population structure, West Point, black sample .

Figure 3. 4. Population structure, West Point, white sample .

Figure 3. 5. Comparison of hh composition, 1880-1910, black hhs, Vinton
Figure 3. 6. Comparison of hh composition, 1880-1910, white hhs, Vinton .

Frgure 3.7. Total persons, by life stage
Figure 3.8. Total family, by life stage
Figure 3. 9. Consumers by hh life stage
Figure 3.10. Workers by hh life stage
Figure 3.11. Dependency ratio, by hh life stage
Figure 4.1. Average number of children born per year of marriage

Figure 4. 2. Average # of children born to married women 50-54 years old

Figure 4. 3. Age-specific fertility rates, 1910, black sample .

Figure 5.1. Agricultural measures by household life cycle, 1880, blacks
Figure 5 ..2 Agricultural measures by household life cycle, 1880, whites
Figure 5.3. Traction by household life cycle, 1880 .

Figure 5.4. Value machinery by household life cycle

Figure 5.5. Value farm by household life cycle .

Figure 5.6. Value produced by household life cycle .

Figure 5.7. Value livestock by household life cycle .

Figure 5.8. Traction by household life cycle and tenure .

Figure 5.9. Percentage that own farm, by household life cycle

LIST OF ABBREVIATIONS

Throughout the presentation of the data, especially in figures and tables, it has been
necessary to use abbreviations. The most common abbreviations are listed below. It
should also be noted that the term "black" has been used consistently throughout the
dissertation. This choice was made on the basis of the general popularity of black,

although other terms, such as African-American or Afro-American, are also appreciated.

blow = black owners
blrc = black renters
blsh = black sharecroppers
whow = white owners
whre = white renters
whsh = white sharecroppers
m =
ur = urban
fam = family
hh = household
0 = consumers
w = workers
dratio = dependency ratio (c/w)
kid(s) = children
ch = children
mo = mother
fa = father
. da = daughter
sis = sister
bro = brother
gr = gm?
fert = fertrlrty
mar(r) = marriage
YT = year

xi

CHAPTER 1
INTRODUCTION

Introduction

This dissertation uses household-level data from a rural community in
northeastern Mississippi to examine interrelationships between tenancy and
demographic patterns. Recently, anthropologists have stressed the need for this kind of
research. One example is W. Handwerkcr. Handwerkcr (1977:259), described
family, fertility, and economics as ”inextrieably intertwined,” but concluded that
”despite intensive investigation, the interdependencies among these phenomena remain
elusive.” Recently David Kertzer (1985: 103) pointed out the continuing lack of
attention to ”the centrality of political economic forces to the understanding of

coresidential processes and household forms”.

More specifically, the study focuses on relationships between household
structure, fertility, and fanning. Studies of fertility and of household composition are
numerous, the latter especially by anthropologists. The household is the setting where
processes of kinship, residence, and domestic production come together. While
family, which extends beyond the bounds of residence, is also a key context for
reproduction, the study of household structure allows many insights into the family
(see Harnmel 1984 or Wilk and Netting 1984 for recent anthropological consideration
of the household).

2

There have been surprisingly few studies focusing on household composition
and formation and other demographic differentials in relation to the organization of
Southem society. This gap in the literature is especially significant because 19th
century Southern cotton production, both under slavery and tenancy, made major
contributions to the ascendency of the United States in a world economy. Katz and
Stern have noted that historical studies of these issues can provide a case study of the
relation between fertility and industrial society ”that has implications for understanding
similar processes at work today all over the world” (1980:229; see also Polgar 1972;
Tilly 1978). Especially relevant is the articulation of semi-proletarianized households

within a capitalistic economy, a prominent phenomenon today worldwide.

Questionnaires!

Much past theorization concerning household structure and fertility falls under
the rubric of modernization theory, and emphasizes the seemingly inevitable impacts of
industrialization. Part of this consensus is demographic transition theory, which
focuses on the shift from high to low mortality and natality (Caldwell 1976; N otestein
1945; Stolnitz 1964). Another part of modemization theory focuses on the nucleation
of Western households (Goode 1963). These modernization theories have become so
entrenched that well-documented empirical challenges have been ignored or rendered
unintelligible. Challenges include 1) stable or rising fertility in some sectors and
declining fertility in others during industrialization, 2) little association between the
timing of industrialimtion and fertility decline, 3) a predominance of non-nuclear
household forms in much of Eastern and Southern Europe, 4) a prominence of nuclear
households in pro-industrial Western Europe, and 5) increasing, not decreasing,
household complexity with industrialization (Hammel 1972; Laslett and Wall 1972;
Kertzer 1984; Tietelbaum 1975; van de Walle and Knodei 1980; Wall 1983).

Dissatisfaction with modernization theory has led many researchers to consider
altermtives that are more sensitive to the structural oppositions of different population
segments, from local to global levels. Anthr0pologists have often been among those
offering objections to the old consensus, on both empirical and theoretical grounds
(Cowgill 1975; Nag 1980; Nardi 1981; Polgar 1971; Raulet 1970), and many of the
clmllenges they have raised are especially appropriate to postbellum rural Southern
fertility. Recent studies of the articulation of prom-industrial or semi-proletarian
households within the market sector (Archetti 1984; Braun 1966; Levine 1977, 1984;
Medick 1976; Smith et a1 1984; Tilly 1978, 1984) have provided an alternative
framework for interpreting high fertility among populations with reduced subsistence
capabilities and increased dependence on early forms of wage labor. Except for the
work of Stern (1983; also Katz and Stern 1980, 1981) on differential fertility in 19th
and early 20th century Erie County, New York, this issue has received little attention
in US. historical demography. However, the comparisons to the American South are
especially appropriate. The cheap labor and regional differentiation (Wallerstein 1983)
of prom-industrialization describes well the role of Southern agricultural production
during the late 19th and early 20th century, when the U. S. was consolidating its
central position within the world economy (Hacker 1970; Palmer 1984). The
important, and some would say dependent (Woodward 1951) position of the American
South within this development has long been noted but rarely approached on the level

of the household or even the community.

Much of the Western experience, especially the decline of fertility during the
later stages of industrial production and organization, are not appropriate as analogies
or predictors for the less develOped waid today. The industrialization of the West

involved complex, interconnected, endogenous processes. Incontrast, many changes

4

occurring in less developed countries today are exogenously induced. Yet certain
phases or settings in the historical experience of the West may hold insights for certain
contemporary settings, especially those involving extremely rapid capital accumulation,
or illumimte some contemporary cases of rising fertility. At the least, more detailed
study of the Western experience should help raise significant questions to apply to

contemporary settings.

Recent studies of prom-industrialization sometimes draw upon concepts like the
corporate nature of the household and the domestic mode of production. The concept
of the domestic mode of production, brought to the fore by Sahlins (1972), following
the work of Chayanov (1966) and Wolf (1966), will be applied to relationships
between Southern tenant households and the larger economy.

In the domestic mode of production, the household’s goals center around
reproducing culturally established patterns of life, with little emphasis on creating or
maximizing economic returns or profits (Sahlins 1972). Work routines and intensities
are tied to the family life cycle, and fluctuate with per capita consumption needs as
household composition changes over time. In the prom-industrial setting, as in that of
contemporary semi-proletarianization, the household-based domestic mode of
production is articulated with a capitalist mode. The extra, non-compensated
contributions of the entire corporate household allow for wages or other forms of
compemation at levels otherwise insufficient for the maintenance and reproduction of
labor, allowing for rapid capital accumulation. These processes of general labor
intensification are usually set in motion by the loss of basic resources like land, which
encourages an increase in wage labor and other forms of market production. Once
dependent upon market production, falling wages or prices for finished goods can be
used to encourage workers to intensify labor, especially incases of little individual

5

control over the means of production. It is expected that data from Vinton farms and
households will show that the composition of households influences the nature of
farming, especially the intensity of farming, as measured in variables such as value of
farm produce per acre or per worker. Especially important should be the number of
persons of different ages, or the ratio of the number of persons in their most productive
years to the number of persons less able to contribute labor, such as the extreme young
and old. This ratio should change as household go through their life cycle. Those
households with less control over their means of production, such as sharecroppers and
renters, may show increased sensitivity to household compositional factors and the
household life cycle. These expectations are based upon the assumption that off-fann
labor is not an important event for most individuals, and that most farms were not able

to generate significant income from means other than farming.

Capitalist production, given its competitive nature, is forced to seek out or
create these types of settings. This is not a unique or temporary feature but rather an
integral aspect of the modem economic system, now institutionalized through part-time
employment, underemployrnent, low wages, and seasonal migration (often rural to
urban) in large portions of the world. In the proto-industrial setting, losses of
traditional resources and the general disruption of societies in place also broke the

constraints of the European late marriage pattern (Hajnal 1982, 1965).

Postbellum Southern society was not strictly a classic proto-industrial settings.
For example, Southern tenancy developed out of chattel slavery, a form of organization
that was relatively rare worldwide in the 19th century. However, Southern society was
characterized by attenuated control of basic resources such as land; general
intemification of labor due to dolls and the laudlord’s control of farming; and
household-based organization of labor, with heavy reliance on the entire family's,

6
including children’s labor. In these characteristics it was highly reminiscent of proto-
industrial and semi-proletarian settings.

H 111 “.5 I1 |

Southern tenancy has often been viewed from the neo-classical perspective of
positive adaptation to market risks. More important, however, are its location within
the wake of a slave-based society, and the negotiations of newly freed blacks for more
autonomous forms of farming. Complete land ownership and autonomy were the
demonstrated goals of many freed blacks, although few managed to secure these goals.
On the other hand, white landowners sought to retain their cultural and economic
position over black laborers while diverting some production risks to them (Flynn
1983; Litwack 1979; Ransom and Sutch 1977; Sholomowitz 1979; Wharton 1965).

The crop-lien credit system upon which Southern tenancy was built was a
compromise between the desires of white landowners and black laborers. Under this
system, the tenant mortgaged the coming year’s crop for the supplies to produce it.
This system provided essentials like seed, tools, and other means (including food) to
laborers who lacked them, and allowed for household-based farming. However, the
tally at year’s end often saw the tenant household in debt to the landlord or furnishing
merchant. In many instances the buildup of debts crushed all hopes of farm ownership
and provided landlords with a lever to persuade dissatisfied laborers to continue or to
increase production.

This leverage was especially important to the extent that many farmers were
operating under a domestic mode of production, regulated more by the inner needs and
rhythms of the household than by market factors. It is assumed in this study that most

.7

households could be characterized as following the domestic mode of production.
Black women and children had largely withdrawn their labor from cotton production
with the end of slavery and seemed hesitant to participate under the wage system; their
labor was regained, however, with the development of household-based tenancy and
the crop lien system of debt peonage (Allen 1975; Cox 1944; Jones 1985).

The choice of the household as the unit of analysis in this study is reinforced by
the fact that newly freed slaves demonstrated their desire for a household-based form of
production, having rejected the less autonomous gang and squad systems preferred by
many landlords. Other studies have suggested that household size and composition
was increasingly important in negotiating the coming year’s contract and could
influence farm size (Davis et a1. 1965; Rosengarten 1974). It is expected that the study
of Vinton will also show a positive relationship between household size and farm size.
This expectation is based upon the assumption that most farm labor was provided by

the household’s members themselves, and not hired labor.

Landlords sometimes demanded set amounts of cotton from their tenants.
larger plots of land per tenant household could determine whether that tenant
household was able to grow small amounts of subsistence crops in addition to their
cotton, and thus reduce their reliance on the landlord or local merchant. As a result,
tenants frequently pushed for larger plots. Conversely, landlords often tried to keep a
tenant’s acreage small, in order to put as many tenants as possible on their land and
farm it more intensively.. In some cases, especially when cotton prices were low,
landlords depended on the interests charged to tenants accounts as much as the profits
from the agricultural produce, a practice called ”farming the tenant.” It is predicted that

this study will show differences in crop-mixes and other strategies according to farm

8

size, household size, and household composition and life cycle, and that better situated

households may grow less cotton and more subsistence crops.

Western!

This study also addresses fertility. Past researchers have documented the
South’s relatively high fertility (Coale and Rivers 1973; Coale and Zelnik 1963; Eblen
1974; Grabill, Kiser, and Whelpton 1958; Okun 1958; Rindfuss 1981). But few
studies have been detailed enough to explain the distinctive patterns observed.
Frequently, researchers assume that the rural nature of the South is sufficient
explanation, without considering societal divisions and dynamies.

Postbcllum black populations have especially been neglected, generally not
included in large-scale national studies until the mid-twentieth century. Consequently,
the decline of black fertility from 1880 to 1940 is not well understood (McFalls and
Masnick 1981). Much earlier work is based on measures such as the child-woman
ratio and cannot address issues like age-specific patterns of childbearing and family

The recent research ofTolnay (1987, 1986, 1985, 1984, 1983, 1981),
especially his analysis of Southern fertility and land tenure in 1900, has begun to refine
our understanding of Southern fertility. This work joins with that of Stern (1985) to
apply more sophisticated methods and a concern for the dynamics of class relationships
to historical U.S. populations. Both Tolnay and Stern (1985) apply the recent findings
of European historical demographers to the United States. Tolnay’s conclusions about
fertility differentials within Southern populations arise from the similarities between
Southem tenancy and the proto-industrial setting. These similarities include a lack of

9
incentives for the European pattern of late marriage given limited pomibilities of land
ownership. Quite the opposite, the household-based organization of labor and the ease
of setting up a new tenant household created incentives for early marriage and
reproduction, lessening the time between generations (Medick 1976). Claims that
mote-industrialization directly increased marital fertility rates are more controversial.
Some researchers suggest an indirect positive effect on fertility, as when intensification
of work or changes in work routines caused a decrease in lactational arnmenorhea (see

Levine 1984, Tilly 1984).

Tolnay's demographic analysis of postbellum Southern populations supports
his prediction regarding earlier marriage for tenants, and supports the comparison with
the prom-industrial setting. Although his analysis showed less clear relationships
between fertility and tenancy, Tolnay concluded, that ”of overriding significance [in
black demographic patterns] was the postbellum adoption of an agricultural
organimtion which emphasized farm tenancy for blacks at the expense of farm
ownership” (Tolnay 1984:306).

This dissermion extends Tolnay's county-level research to the household-level.
It is predicted tlnt land tenure will be an important variable in explaining fertility
differences. More specifically, it is predicted that tenants will show an earlier age at
marriage, and a slightly higher level of fertility. Although detailed analysis of age at
marriage can only be carried out for one data set, 1910, it is predicted that there will be
some indication of a general drop in age at marriage, perhaps through changes in
household compositional. This expectation is based upon the assumption that tenancy
increased in the study area over time. It is also assumed in this study that fertility is
natural, or nearly natural, especially in the rural sample, and that these differences in
fertility stern largely from increased exposure and not an increased pace of fertility

10

within marmge, according to tenure. It is predicted that these differences will hold for
both the black and white samples.

Way

The study area is the rural community of Vinton, Mississippi, which consisted of
dispersed houses and farms and several country stores and churches. A few doctors and
storekeepers in the white sample were the only occupations other than farming in Vinton.
The temporal focus is 1880 to 1910. The beginning date of 1880 is sufficiently removed
from emancipation to insure that the tenancy system was well-developed. The end date of
1910 is early enough to minimize changes associated with large scale rural to urban
migration, which stepped up considerably after WWI, industrialization, or agricultural
mechanimtion.

Data for this smdy consists of a variety of documentary records collected by the
author. The main types of documents used are land deeds, chattel deeds, estate records,
population and agricultural schedules from the federal census, and a private account ledger.
Some of these data were collected in 197 9 and 1980, as part of Michigan State University’s
Tombigbee Historic Townsites Project. This project used archaeology, oral history, and
archival research to study three communities in Clay County, Mississippi (Cleland and
McBride 1983; McClurken and Anderson 1981). Two of these communities were
nucleated town which existed in the mid-19th century. The third community was a more
dispersed community, which existed from the mid-19th century into the 20th century. It
exists today as a dispersed locality or neighborhood, although most service functions, such
as country stores, a ferry, or a post office, do not exist. Land deeds, chattel deeds,

census records, and a store account book from the Vinton community were the major data

11

sources for this study. Besides the data collected during the Tombigbee Historic Townsites
Project, additional data were collected by the author in from 1981 to 1985. These data
included additional land and chattel deeds, court cases, and additional census schedules.
Copies of census schedules were purchased from the National Archives, and other
materials were studied and, when necessary, photocopied, from the Clay, Monroe, and
Lowndes County, Mississippi courthouses. The private ledger was photocopied, with the
permission of its owner, before it was donated to Mississippi State University Special

Collections.

The analysis was accomplished by use of a variety of methods. General historical
research and reconstruction of landholding patterns and tenant-landlord relationships was
used to provide a historical narrative of the study area. This narrative establishes the
nature of tenancy within the study area Subsequent analysis on the household was
conducted by analyzing the composition of households, their life cycle stage, and the
position of persons within households. The latter two approaches are important to mediate
the generally static nature of household compositional analysis. Analysis of marriage
patterns and fertility was based upon computation of a series of measures, some standard
and some experimental. Specific methodologies are described in their corresponding
clmpters. These include calculation of age-specific fertility rates, indices of the level and
character of fertility, and singulate mean age at marriage. A major thrust of the analysis
was to calculate these measures separately for subsarnples broken down by race, residence
(rural versus urban), and tenure. Analysis of agricultural was accomplished by summary
statistics on agricultural holdings and production, also broken down by owners, renters,
and sharecroppers. Simple correlation (Pearson’s R) analysis provided assistance in
examining relationships between the agricultural and demographic data.

12

Most of the numerically oriented analyses were facilitated by construction of a
series of linked databases for 1880 and 1910. The most basic databases were person and
household databases, constnrcted from the federal census. The census is organized by
districts, with each district having its own (usually locally known) enumerator. For 1880,
each database consists of the entirety of two census enumeration districts. The first district
used was the northern half of Beat One (a local county subdivision) of Clay County. This
encompmses the Vinton community. The southern half of Beat One, which was
enumerated as a separate district, was not used because it was composed predominantly of
the Waverly community. The second enumeration district used to in the 1880 data set was
the Darracott enumeration district of Monroe County, which bordered Vinton on the
northern side. Historical research established the close interaction of the Vinton and
Darracott communities during this period (McClurken and Anderson 1981; McBride 1983).
Use of these two census enumeration districts resulted in a total data set of 1,184
households from the population schedule of the census. One ward (108 households) from
West Point, the county seat town and the town closest to and most used by the Vinton
population, was also entered into the person and household databases for comparison in the
fertility analysis.

Each person from each households listed in the census was entered as a separate
record into the person database. This record described characteristics of the person, such
as age, sex, race, occupation, and so forth. Each household was then entered as one
record in the household database. The household record described characteristic of each
household, such as name of the household head, household type, life cycle stage, tenure,
number of persons, number of family members, types of non-nuclear family members,
number of non-relatives, and so forth. Each person in the person database was assigned a
household number, according to the household in which they resided. This same

l 3
household number was assigned to the corresponding household in the household
database, and provided the link between the person and household databases.

Both the household and person databases were constructed from the population
schedule of the census. The next step was to link the household database with information
from the 1880 agricultural schedule of the census. This schedule lists each farm by the
name of a farm operator, along with an variety of categories of information on agricultural
holdings and production. This agricultural information includes whether the farm was
owned, worked on shares, or rented for cash; the number of improved acres, the number
of unimproved acres; value of livestock, value of farm implements and machinery; wages
paid to hired labor; counts of livestock, by types, and value of same; acres planted to
individual crops; and yields of individual crops.

The farm operator was the person considered to be in charge of the farm, such as
the owner, or in the case of a tenant farm, the head of the tenant household which rented
the farm. Almost without exception, this person was the head of the rural household, as
given in the population schedule. These names of the farm operators were entered into the
agricultural databme. Once sorted alphabetically, the names of these operators were
matched to the names in the person database, and through the household number, to the
household database. Any farms that could not be matched to an entry in the household
database were not utilized. A total of 467 households from the household database were
matched with entries from the agricultural schedule. Very few households in the
agricultural schedule could not be matched to the household database, but a large portion of
households in the household database were not listed in the agricultural schedule. This is
because their members worked as agricultural day laborers and did not own or rent their
own farm. Thus is should be remembered that the agricultural analysis, or any analysis

14

that is broken down by tenure, deals only with those persons who operated their own
farms, whether they sharecropped, rented or owned the farm.

Person and household databases were constructed for 1910 in the same manner as
described above for 1880. The 1910 Vinton sample is composed of all of the northern half
of Beat One, or 686 rural households. While most of the detailed analyses focus on the
rural Vinton community, Wards One and Four (525 households) from West Point were
also entered into the person and household databases. This sample is roughly the northern
half of West Point and encompasses large white, black, and mixed neighborhoods. It is
used predominantly in the analysis of fertility, because of the importance of rural and urban
differences in fertility.

Like the 1880 census, the 1910 census provides usual demographic data such as
age, sex, and occupation, which was entered into the person database. However, the 1910
census also provided more detailed information on marital status -- including whether a
marriage was a first, second, third, or, in a few cases, fourth marriage; and the number of
years in the present maniage, which was also entered into the person database.
Demographic information also included the number of children born to each woman and
number of children surviving, along with a listing of these children and their ages, if they
resided in the household. This also was entered into the database.

Unfortunately, the manuscript agricultural schedules for 1910 (and 1900), which
appears to have given tremendous detail on tenancy arrangements, have been destroyed by
act of Congress. However, the 1910 population schedule included the crucial variable of
whether the household’s home or farm was rented or owned. This information was
entered into the household database. Some additional agricultural data were available for
the 1910 time period from a private account ledger from a Vinton commissary. This ledger

l 5
gives a detailed accounting of the supplies furnished to a group of tenants, as well as other
charges, like rent, and credits, such as pounds of cotton produced. Although this ledger
was for the year 1911, it was linked to the 1910 household database with the assumption
that most households would not have undergone significant changes over one year. A final
database was constructed for the 1910 fertility analysis. This database listed all children,
by the age of their mother, from the 1910 person database.

Atmnsnnentatghanters

Chapter 2. discusses the history of tenant farming in the postbellum South and
gives a brief narrative history of the Vinton community. It highlights the major theoretical
issues in the study of tenancy which the present study can address, and traces the
development of tenancy in the study areas. Chapter 3. presents the data and analyses on
household composition and the household life cycle. It presents a look at household
composition from the perspective of the household unit and the individual. Chapter 4
presents the analysis of marriage and fertility. Chapter 5. presents the agricultural data and
amlysis and discusses relationships between the agricultural data and demographic
stnrcture and processes. This discussion includes consideration of Clmyanovian
relationships. Chapter 6 summarizes major findings from the study, presents comparative
data from select studies, and assesses the contributions of the study. The Appendix
presents the various systems of classification or categorization used in the analysis of

household structure and composition.

CHAPTER 2
THE CONTEXT OF SOUTHERN TENAN CY

lgtrmlugtign to Southern Tenangz

This chapter provides the historical setting for the analysis of household
composition, fertility, and agricultural production. It introduces the reader to postbellum
Southern tenancy in general, and discusses historical developments within the Vinton
community. In the latter task it draws heavily on McBride (1983), with additional use of
chattel deeds and court depositions from the Clay County Courthouse, West Point,

Mississippi.

Agricultural tenancy was well-established in the South before the Civil War, as well
as in other regions. A number of factors encouraged tenant farming, including tight
money policies and frequent foreclosure by powerful financial institutions, the speculative
nature of American agriculture, which drove up land prices and made short term leases
profitable (Goldcnweiser and Truesdell 1924). Antebellum tenancy has frequently been
interpreted as an ”agricultural ladder” that allowed those initially without the means to buy
land and supplies to engage in productive farm labor and accumulate the means to become
farm owners . After the Civil War, agricultural tenancy became very common in the South,
especially those areas that specialized in cotton. This tenancy rose to such an extent that

the older interpretations of tenancy were no longer viable (Cox 1944).

16

17

Tenancy in the South after the Civil War was frequently the result of the
subdivision of a large plantation into a number of tenant farms. By 1910, fifty percent of
all farms in Mississippi were tenant farms on tenant plantations, higher than any other state.
Another 20 percent of all farms were tenant farms not on plantations, usually single tenant
farms. These tenant "plantations," as they were frequently called, could be composed of
flora one to 50 smaller individual farms. Of these tenant plantations 67 percent were

composed of five to 10 tenant farms (U .S. Census 1913:884, 1916).

By 1930 72% of all Mississippi farm operators were tenants, and 47% of the entire
Mississipi population were members of tenant farming households. This figures compare
to national figures of 25% of all farm operators in 1880 and 42% of all farm operators in
1930. Nationally, there were always more white than black tenants, although blacks might
individually have a higher rate of tenancy. However, in Mississippi, black tenants greatly

outnumbered white tenants (Cooper 1933).

Tenant farming was especially concentrated in cotton areas. A special report on
cotton production in the United States in 1880 (U .8. Census 1884) showed that 43 percent
of the tilled land in Mississippi was devoted to cotton, more than any other state. One of
the most importanf cotton regions of the state was the black prairie, often called the black
belt because of its narrow formation that swept through the northeastern and northcentral
portion of the state. The richness of the black prairie soils had attracted some of the
earliest settlers to Mississippi. Figure 2.1 (taken from U.S. Census 1884) shows the
location of the study area within the black belt. Clay County, within which the study area
is located, devoted 51% of its acreage to cotton in 1880. This figure is lowered by the non-
cotton lands in western Clay County. On the eastern side, where the study area is located,
even higher proportions of land were planted in cotton. Although cotton prices fluctuated

during the thirty years encompasses by the present study, being especially low in the

18

1890s, major changes in cotton agriculture did not occur until the drop in demand
associated with WWI (Daniel 1984). Bondurant and Welch (1956) suggest that cotton
prices from 1870 to 1915 were relatively stable and followed closely the wholesale price
index of all commodities. Black populations usually did not begin to show the effects of
large scale out-migration until the next decade, when the growth of jobs in northern
industries and the boll weevil, and dissatisfaction with the increasingly extractive policies
of landlords encouraged abandonment of tenant farming (Fite 1984; Higgs 1984; Kirby
1983).

Thus tenant farming was a proven means of organizing agricultural production in
the United States, and one that would eventually characterize much of the South. A large
percentage of ex-slaves had worked in agriculture during slavery, and wanted to continue
farming. Although most err-slaves demonstrated a desire to farm for themselves, mom
lacked the means to buy or rent land or to even acquire basic agricultural tools and supplies.
Although reconstruction policies initially called for widespread distribution of land to ex-
slaves, such policies were changed, not fully implemented, or repealed. Ex slaves lack of
land, or credit resulted in a massive disequilibrium between labor and land in the South
following the Civil War.

Tenant farming, although always a possibility, was not the initial solution tried in
most part of the South following emancipation. Ex-slaves themselves were not accustomed
to this method of farming, and neither was the majority of Southern white landowners.
Many landowners probably had difficulty imagining persons they had once owned as
slaves, and for whom they had elaborated a massive ideology of paternalism and
dependency, farming on their own, or were generally reluctant to give up any control of
agriculture on their lands. Instead, most early attempts at reorganizing agricultural
production were based upon wage labor. This was the system formally promoted by the

£3 ace—a £55 8.3 mean we .8580.— ._.~ 95mm

 

19

LI 1‘ .3 Jo rel I H
. . .
Mex. . 111111111... _
. 1 $159.45.: . ”his-Juan“ .
n .e l .g.i_=..s.euu
, ac nchuEnEEEgc 1111.1 .n I
4.55.1.7; Huh
3?... . . an e. a” I
ZDPHO; .fi. unwawchth 455.. h
ds.—anus»!— . . .2.! a. I n
meatjuzgénewmgubofiéem . . 21. I m
mmzkm Pawn—Damn VOHHOU . .n .
. 3 H _
r 11).... _ 1 1. B n
\I li/H/flfwlfll rlJ ..
///«/.vl..lnM! late 7. I _ P.
ant. ,Nrtumu .
,.. lrlrwmm. PS...“ .1
. .. . e
. .. 7.... _
u L}. (’Lun
. , . fit. {/1
n . .. Jr if, -1: . V?
m. ._ a. let» l.
M, .. .I z 3 __
. llr .
. . ll _
u .u . vi! 1/ «Jill
Wu 7...“! __
. ?r _
1.. . , 1- _
. . A _
l . ., _ (r- r _
- r .. r i w Fifi/>1 .1 . r»
u . r . w . .. 1.13.2941, xx (.1 ._ .2
. rum... rti, . 1.1/1..- -1 /. .r ._
n . .. ac... /. rimmuvff frettinzfnkl .
_ MW. an. .. . . . .. \
. 2 TM . .x l. )1 l t ,
9 - a, ,. ._ .1 . \ Jaw/71.x,
t. t z. ; . >131... . . u
. f if .1, r _

 

.J s....l..1.
. pears—zszker

 

 

                  

 

      

 

 

 

 

 

 

 

 

f , . «trilllh [Till i ,
I 1V x . .. .. \ltlllr\\ .llfi
. .— \\ r\. v / .. l/ . 11\\.I \. .. w—Jn
.r, .111? 1. / J , / L llhlJl .
. . .. . .\ \. e“— / ., ,, m _ .. . ,. ( _
. SKIN l a a .r. . ,. , .., .. 1..., l 1- 1.. 8)).
. . m o 1 \l _ I I I
, . . _ .. . . / .1... n l . .
I... . .. > , . if ,...._.. L -t/. (11.1,
"rears. K, . .2 .3. s . ., Le. .I L. . ...s.. .2 .3 .5 .8 .3” .3.
II: t; Ill

 

20

Freedmen’s Bureau. Often landowners tried to organize blacks to work in gangs, much as
they had under slavery. Arrangements as to provisioning and living conditions, the level
and method of payment, supervision, and other factors varied greatly.

There were many problems with this system. A true test of wage labor was never
possible because landowners were unable to pay wages sufficient to support black
laborers, on a regular basis. The shortage of specie and credit that characterized the South
after the Civil War severely limited payment. As a result, landowners often could not pay
until the end of the season, during which time they had to provide for the physical
necessities of their laborers. To many parties, on both sides of the transaction, this
arrangement was too similar to the situation under slavery. A cotton agent on the Sea
Island noted that , ”one thing the people are universally opposed to. They all swear that
they will not work in a gang, i.e., all working the whole, and all sharing alike” (Pearson
1969). Black laborers’ general dissatisfaction with wage work can also be seen in their
demands for converting the fixed wage to a wage determined as a share of the crop. This is
despite the fact that a fixed wage was more secure, and in some circumstances, more
lucrative (Shlomowitz 1984). It is not surprising that blacks would have been dissatisfied
with wage labor in any form, since most had expected their own farms once they were

freed.

Also, black laborers expected wages sufficiently high to support not only
themselves but also non-working members of their family, a situation that many
landowners/employers would not support. A frequent complaint of landowners/employers
during the early years after emancipation was that many black women and children had
withdrawn from agricultural production. For example, contemporaries and modem
economic historians estimated that the labor devoted to agricultural production during the
late 18605 and early 1870s was only about one-third of pre-emancipation levels, in large

21

part due to the withdrawal of women and children (Gutman 1976; Jones 1985: Kolchin;
1972; Litwack 1979; Ransom and Sutch 1977; Shlomowitz 1984).

This situation was viewed with disapproval not only by Southerners but by many
of the northern abolitionists who had strongly opposed slavery, especially since women
and children working for wages was a common occurrence in Northern factories. Jones
(1985) suggests that northemers expected blacks to pursue wage work in a self-interested
fashion, and that they misinterpreted black women’s lack of desire to do so as attempts to
imitate middle-class white norms. In these assumptions, the northern critics disregarded
the strong family ties that bound black households, underestimated the desires of black men
and black women to restrict the contact and power white men had over black women, and

underestimated the desires of black women to work directly with and for their families.

It can be argued that as long as black women pooled their wages into a family fund,
they were working for their families. This case has been described as the family wage
economy by Tilly and Scott (1978), among others, as characterizing much of white
women’s wage work in the 19th century. However, the lower wages set for women by the
Freedmen’s Bureau, and the use of the gang organization, may have greatly decreased the
incentive for black women to continue a form or labor which was repugnant and which
separated them from their children, and often husbands, for most of each day. This
interpretation is supported by the fact that more widowed and single black women
participated in wage labor (Flynn 1983; Jones 1985). Woodman (1985) and Strickland
(1985) also argues that contemporary critics (and subsequent scholars) underestimated the
extent to which blacks objected to wage labor itself, regardless of the level of wages or
working conditions, as violating their own moral economy.

22

The squad organization of labor was probably the first adjustment to the low
success of wage labor by gangs. In this form, groups of laborers, usually 5 to 10, would
work together. One member would have ultimate authority and responsibility for the work
of all. Squads were often groups of related persons, in some extremes one nuclear family,
and in this way they foreshadowed household-based sharecropping. Although less
objectionable than the gang system, the squad system was not totally successful either.
One problem which was never resolved was how to insure an equitable division of the
product among the various laborers in the squad, matching the retum to the varying levels
of work contributed. Squads also did not meet the desire to work as family units

(Shlomowitz 1979).

Sharecropping on separate parcels, farmed by family units, was tried in some areas
in the late 1860s. It aroused bitter complaints from the majority of landowners, who were
still in favor of the gang wage system (Wiener 1978). Wright (1986) estimates that it was
not until 1880 that the individual sharecropping system had more fully replaced the wage
labor system, although of course some scattered wage labor continued throughout the entire
postbellum period.

The system of sharecropping that arose in the South was in some ways similar to
wage labor, since most families brought only their labor to the farm, with all supplies being
fitmished by the landlord. The title to the crop actually remained with the landowner, with
the sharecropper being paid out of the returns from the crop. However, sharecropping
differed from the old gang system and from more typical wage labor in crucial ways. The
household replaced the gang, and the constant supervision of an overseer was replaced by
intermittent visits by a landlord or his agent. Individual tenant houses were usually
constructed on the farm, as opposed to the centralized housing that characterized gang wage

2 3
labor (often the old slave quarters). And renters, as opposed to sharecroppers, did have a
stronger legal claim to the crop.

Although many landlords were skeptical of sharecropping and renting, they were
largely formd by the persistence of blacks to adopt this organization. Many landowners
were especially against renting land to blacks, since renting usually assured the tenant
greater autonomy as well as legal title to the crop. Many white landowners were skeptical
that blacks would provide the level of labor the landowner desired without coercive
pressures (Fight 1984; Flynn 1983; Hahn 1983; Wiener 1978). However, landowners
could also see positive elements in sharecropping and renting, among these being the
opportunity to cut down supervision time and costs, the sharing of standard agricultural
risks with laborers, and the hope that the laborers would increase their interest and devotion

to crops when farming as a family unit and sharing the produce.

Most scholars agree that the entire black family was drawn back into cotton
production under sharecropping, and renting. The variety of sharecropping and renting
arrangements was great. A study in 1910 concluded that ”the details of these variations in
dealings of landlord and tenant are practically endless” (U. S. Industrial Commission 1901,
cited from Alston and Higgs 1982). The portion of the crop that went to the landlord
versus the tenant varied both regionally, and on a case by case basis. The crop division
usually bore some relationship to the relative inputs supplied by each party. True renters,
whether the rent was paid in cash or crops (often called ”standing rent”) had a stronger
legal claim to their holding and crop. Sharecroppers could also be split into true
sharecroppers, who provided only labor, and share tenants, who provided power,
equipment, and a portion of the fertilizer and other inputs. One specific form of share
tenancy was called ”thirds and fourths.” The tenant furnished provisions, tools, and the

team but only part of other inputs, and received three-fourths of the cotton and two-thirds

24

of the corn as payment (Taylor 1943; Welch 1943). This was a very common system in
Mississippi, and it is likely that a large portion of those farmers reported as sharecroppers

in the federal census were in fact share tenants.

More complication arrangements sometimes included direct contributions of labor
from the tenant farmer to the landlord, to be applied to other farms tlmn the sharecroppers’.
For example, Lewis (1984) reports an arrangement in South Carolina where one day of
labor per week constituted the sole rent. This is a variant of the ”sharing of time”
agreement discussed but shown to be rare by Shlomowitz (1979). Also, one plantation
might use several different arrangements, depending on the needs and resources of various
farmers, and the level of input and supervision provided by the landlord varied
considerably. Many additional or side arrangements were often made to a basic contract,
as for example when tenants were paid wages for long-term maintenance work not
considered part of their normal responsibility to their crop. Tenants on a plantation where
the landlord lived elsewhere often had the most autonomy, although the landlord may have
employed an agent. Renters generally had more autonomy, although Edwards (1913)
suggest that this was not originally the case (Edwards 1913; Woofter 1969). Mandle
(1978) points out, however, that racial prejudices restricted the autonomy of all black

farmers, even renters and owners.

Renting usually required ownership of at lesst one mule and other personal
property. Nationally, the proportion of tenants that rented instead of sharecropped
increased until 1900. Although this trend is consistent with the ”agricultural ladder”
interpretation of tenancy (DeCanio 1974; Higgs 1977), this trend has also been linked to
the transfer of lands from local landowners to merchants and factors, many absentee, who
preferred the lower involvement and supervision of renting (Fite 1984). After 1900, the
proportion of sharecroppers increased (Alston and Higgs 1982). This trend is explained by

2 5
Wright (1979) by increases in the prices ofmules from 1899 and 1918. It may also be
related to reform policies on the part of landowners to increase profits from tenant
plantations during this period. Such policies were explicitly stated and enforced on a
nearby plantation in Monroe County from 1901 to the early 1920s. The new policies
resulted in the removal of most renters and their replacement with sharecroppers, who
could be controlled to a greater degree by the landlord (Kern et al. 1982).

Ihflintmficmmunitx

During the postbellum period, Vinton was a loosely dispersed rural community
along the Tombigbee River, in Clay County, Mississippi (Figure 2.2). Exact community
boundaries are difficult to draw (Cleland 1983). The population was predominantly black,
depended on cotton production, and tenant farmers. In the antebellum period, the area had
been characterized by cotton farms and plantations, but also included the nucleated riverport
communities of Colbert (1835- 1847) and Barton (l848-circa 1865), and the semi-nucleated
crossroads community of Vinton (l848-present). Settlers in this area had frequently gone
to the Lowndes County seat of Columbus for business, social activities, and to acquire
merchandise, although those in the northern part of the area also patronized the Monroe
County seat of Aberdeen. Many of the large planters also relied on cotton factors from

Mobile to market their cotton and procure supplies.

The nucleated town of Barton began to decline in the mid 1850s when the Mobile
and Ohio Railroad bypassed it for the town of West Point, about 10 miles to the west.
During the 18608 and 18708, many of the white residents, who in large part had lived in
Barton (Figure 2.2), left the area for West Point and other places. For example, of the 28

white families known to have lived in Barton or near the Vinton store in 1860, only 10

26

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

Lumen

 

 

 

 

/\-/ 13.3 ”flak-(LVl {ll->4 /-.\|\
w on. .
... a $.35): a
.. .. i. (l a
_
w _
a uuuuuuuuu L .
a IIIIIIII . .. r lllllllllllllllll
Hill-r “- .1 a
_ "- _ .
r."- llllllllllllll IC. ... lllllllll m W
. n .r a M s
w w. w _ v FIIJ S W
L . .I
M m. u w. _ m m m w
_ .1 ((((( Lutlr m. .. M
e a . m m

 

--— CW MI

Figure 2.2. Location of study area and nearby communities.

27

were located in the 1870 census. Few new white families are known to have moved into
the area during the postbellum period, and the variety of occupations represented decreased
dramatically. Some of the white families who remained in the area experienced financial
difficulties during the postbellum period, as was typical in most parts of the rural South.

Since slave counts are not known for the entire study area, it is hard to say how
many ex-slaves used their freedom to migrate to other places. The frequency of sumames
such as Matthews, Fields, Strong, Keaton, Lloyd, or Cox among the postbellum black
population correspond to those of the larger antebellum planters in the area, and suggest
that many ex-slaves remained. The extent to which Vinton residents were involved in the
Reconstruction government and activities is difficult to determine. There are records of
local complaints to the Freedman’s Bureau concerning leases or contracts now on file at the
Natioml Archives. Letters of the Shaw family of Darracott, just north of Vinton, suggest
that many white residents also experienced considerable difficulty in adjusting to the
changed social and economic conditions.

Monroe County, into which the northern portion of Vinton extended, was an active
Ku Klux Klan area. One of the largest landowners south of Vinton was Thomas Martin,
an absentee landlord from Tennessee who was influential in the establishment of the Ku
Klux Klan. The fall of 1875 was so full of ”political excitement,” as described by one local
resident, that the cotton crop was not completely gathered (Clay County Case File 236). In
November of tlnt year a group of l l of the largest landowners from District One of Clay
Cormty, many of them from Vinton, formed a committee and put out notice that they would
refuse to rent to anyone who voted the Republican ticket. They also warned that they
would refuse to rent or deal with anyone associated with such characters, and that a penalty
of a double rent would be inflicted on tenants who did. They specified that they considered

anyone who refirsed to cooperate with their committee unworthy of public confidence and

28

trust, and that the names of such persons would be sent to the local papers. The
democrats were back in office in 1876.

Little information is available on social activities and networks during the
postbellum period. A number of local clubs, like the Masonic Lodge, the Grange, the
Vinton Gun Club, and a reading club, were organized by the white population. Political
speeches, rallies, and picnics were also popular, as were seances at the Shaw home in
Darracott. The strength of rural community bonds is apparent from the Vinton column in
the West Point paper, where one resident’s illness or accident became the concern of many,
or a courtship or mishap might be the occasion for humor. Eva Coletrane, a 15 year old
white resident, described the Vinton community in 1887 in this way:

Vinton is about ten miles from West Point, two and one half miles north of
east. What a great blessing a post office is, even without the advantage and
privilege of a town or city. There are so few white people living in our
section that living here borders on confinement. It might suit one who does
not like to be crowded, but one of my age, with a strong desire to see a great
deal of the world, suffers a little sometimes, like the inmates of a cage. Were
I qualified to write an interesting letter it would not be a batch of locals. We
have no landings in use, no depots, and but little visiting; hence the
advantages of these things furnishes no material for a Vinton correspondent.

I do not mean to complain, but merely to apologize for having very little in the
way of local items to write (West Egint Leader, 24 February 1887).

Even less in known of social activities and networks in the black community. Oral
history has established the importance of the local churches and schools, and of a strong
mutual aid network among black residents (McClurken and Anderson 1981).

Freed blacks demonstrated strong desires for schools for their children (Gutrnan
1976). County warrant books recording payments to teachers or for repairs to schools
indicate that schools were established for black students at Vinton in the 18703. Some of
the early schools within the study area were held in churches. These include the schools

29

held at Free Grace Baptist church, the Concord churches, and Paines Chapel west of
Vinton. A separate school, the Andrews School, was established near the Bethel Baptist
Church.

Schools, both black and white, were continually being moved around, presumably
in an effort to minimize walking distances and spread out enrollments. In 1891, a new
school was established in the southern portion of Section 35 (see Figure 2.3 below). The
main school for black children at this time was on Town Creek, in Section 34, with the
Fields School serving the southern end of the neighborhood. The white Vinton school
usually had only one instructor at a time, while the main black school at Town Creek
usually had from two to four teachers or assistants, in response to the larger number of
black children.

School was in session for 80 to 100 days a year during most of this period, from
November to March. Attendance was undoubtedly affected by the necessity of children
helping their parents in agricultural chores. The Vinton correspondent to the West Point
paper noted in March 1887 that ”the school will be lessened in number soon, as some of
the pupils will have to discontinue. The busy time of the farmers is near at hand.” A state
school census in 1880 indicated that attendance at the black Town Creek school in 1880
was 76 students, 29 males and 47 females.

Attendance at the white Vinton school in 1880 was 30 students, 13 males and 17
females. The fact that the number of white students was almost 40 percent of the number
of black students suggests a much higher attendance rate for the white sample. School
records from the 1890s suggest that total attendance of black students was usually from 85
to 100 pupils during the 1890s. The male to female ratio was usually nearly equal, with
females often slightly more numerous. Enrollment at the white Vinton school ranged from

3 0
33 in 1889 to 13 in 1900. At the beginning of the decade there were many more male
students, but by the turn of the century equal numbers of boys and girls attended school.
Because of the fluctuations in school district boundaries, it is hard to estimate the total
school attendance for the wider Vinton neighborhood used in Chapters 3-5. The children
of Vinton merchant W. E. Trotter attended private schools in other communities. Trotter

was likely one of the wealthiest landowners in Vinton, and most Vinton residents probably

could not afford private schools.

The town of Barton, located about one mile south of Vinton, had been severely
damaged by the coming of the railroad to West Point in the mid 1850s and ceased to exist
during the Civil War. The community of Vinton, which had to some degree been
overshadowed by Barton during the antebellum period, now expanded. The center of the
Vinton community remained the Vinton store complex. It included a house, store, post
office, ferry, blacksmith shop, and cotton gin (Way and McBride 1983). However, in
1876 the Vinton school and church, which had been located near the Vinton store, were
moved several miles north to better service the entire rural Vinton community. The church
was eventually merged with a church from the neighboring community of Darracott, and
called Bethel Baptist. It was predominantly composed of white residents, although its
membership roles for 1878 to 1900 include two blacks. Most blacks attended the Concord
1 and II churches, located in the western portion of the community; the London Chapel or
Town Creek churches, located in the southern end of Vinton; or Free Grace Missionary
church, organized in 1869 and located near the Bethel Baptist Church in the northern
portion of Vinton.

One of the most important social and economic institutions within the community
was the Vinton store. During the antebellum period, the Vinton store was owned and run

by several persons. However, the last and main owner was William E. Trotter. Trotter

3 1
had spent some of his young adult life in nearby Columbus, Mississippi, before
establishing a store in Moscow, Marion County, Alabama. Shortly before his departure
from Marion in 1854, R. G. Dun and Co. agents reported that he was the best businessman
in the county and that he had made a small fortune in trade. During his stay in Moscow,
he married Sarah A. Moore, who was probably related to the Moore family who owned a
large plantation near Vinton. William E. and Sarah Moore Trotter moved to Vinton, and in
1855 Trotter joined the mercantile business of his brother-in-law William H. Moore, at

Vinton. Moore sold out to Trotter within several years.

Under Trotter‘s proprietorship, the Vinton store prospered. Trotter’s land
holdings during the antebellum period had been concentrated around the store, as most of
the prairie land to the west was owned by large planters. However, during the second half
of the nineteenth century, Trotter’s holdings grew tremendously. Beginning in 1870, he
bought much of the land just west, north, and south of his home and store. He also
acquired several parcels on the east side of the Tombigbee River, within several miles of
the Vinton store. In 1880, he added another section to the western holdings.

Figure 2.3 shows Trotter’s lands, and gives dates of acquisition and resale or
mortgage. He acquired several parcels through mortgages. It is likely that other
transactions, although not specified, may have represented settlements of debts owed to his
store. By the late 1880s Trotter owned at least 4,369 acres, more than most of the largest
antebellum planters in the area. For example, in 1860, only 2.2% of all plantations in
Lowndes County were over 2,000 acres (Stephen McBride, personal communication
1990). This process of concentration of holdings during the early postbellum period has
also been demonstrated in Alabama and Georgia (Hahn 1983; Wiener 1978).

3 2

Acquisition of land by a mral furnishing merchant like W. E. Trotter was not
unusual. The demise of the factorage system which had supplied much of the rural South
created a large gap in finance and mercantilism. These problems were increased by the
scarcity of banks and specie in the South, a lack of alternative credit arrangements, and a
lack of capital for security. Land values after the Civil War were low, reducing the
collateral of owners, and then a large percentage of the population, the freed slaves, owned
no land and little personal property. The rural fumishing merchant and the crop lien system
solved these problems (Ransom and Sutch 1977; Woodman 1968). With credit from a
northern wholesaler, the merchant sold supplies, usually on credit, to local residents.
Security was provided by a lien on the coming year’s crop. Lien laws were established
soon after the end of the Civil War. They were described by Southern historian C. Vann
Woodward (1951: 180) as ”one of the strangest contractual relationships in the history of
finance.” Yet Woodman (1968) points out that in some regards they were a continuation
of the factorage system, under which goods were advanced on credit to a planter who
promised to let the factor handle the coming cr0p. While most antebellum transactions
were conducted by word of honor, the postbellum arrangements were more formal, with a
lien recorded on the crop. Also, the antebellum factor was usually no more than an agent
for the planter, while the postbellum merchant took actual possession of the crop through
the lien.

The furnishing merchant could use a local knowledge of farmers to estimate their
solvency and productive capacity in a way that outside suppliers could not. This
knowledge and the dependence of all farmers, not just tenants, on credit, led the rural
merchant to establish territorial monopolies with virtually no competition. The local
fitmishing merchant often not only provided supplies, but managed the marketing of the
agricultural crop for local farmers, served as a bank in dispensing cash, served as the legal
intermediary for persons (especially blacks) unfamiliar with legal and judicial systems, and

33

52:8 52 oE— :28: .m .3 he 955$:an wee.— .m.~ 0.5me

 

 

wnm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

onm um
32....
39.5
m8;
mmm nnm
2.2m
9.9...
aged 05....
mum own
«8....
89.5
2.2...
\ Em nwm
.. mom-E mom-.6
. 22.0 new;
1 . $2.2
a mom; 89.5 km. .o
a .o . new. o
2.9 moon;

 

 

 

 

 

Bantu

cowmwtcE I E

 

poem—Eon n a

:33 3:! o 2:30.: IIW

 

 

"HM-II.”
I. 0
mm m; m. m
#095
mom. .0
mmwfiomm
boo. E
.e

 

 

nmw; QNm

5mm. .5
0mm. .0

 

 

3 4
generally became one of the most central figures in the rural community (Clark 1944, 1946;
Ransom and Sutch 1977).

Supplies were usually sold on account, with higher prices for credit purchases, and
interest at 10 to 20 percent of the account. Local Vinton merchant W. E. Trotter charged 10
percent interest, as did a merchant in nearby Pickens County, Alabama (Wesson 1980).
Ransom and Sutch (1977) suggest from a study of a number of merchants in the South that
total interest and credit charges averaged over 50 percent, in addition to wholesale to retail
mark ups. They also suggest that many merchants, as well as landlords, employed a
number of devices, including faulty bookkeeping, intimidation, and violence to keep black
laborers in debt, and thus working. However, these methods are not necessarily in
evidence when a merchant showed a profit, since the markup and interest alone should

have created favorable business conditions.

Since landowners also desired a lien on the coming crop, there was a built-in
conflict between merchants and landlords. Many states passed legislation to help resolve
this conflict. In Mississippi, for example, the landlord lien was given priority over that of
the merchant (State of Mississippi 1880). Other states passed different legislation,
although most states within large plantation regions followed this path.

The conflict between merchants and landlords was sometimes resolved by merging
these two functions in one individual. There is some debate whether the dominant pattern
was for former landowners to become merchants, or merchants to acquire land (Hahn
1983; Ransom and Sutch 1977; Wiener 1978) In the former case, a landlord would often
begin by setting up a commissary, mainly to supply tenants, and later expand into general
merchandising. In the latter, a merchant would acquire land, usually surrounding the
store. Wiener ( 1978) suggests from a study of a number of merchants in Alabama that in

3 5
plantation regions, planters usually became planter-merchants, and remained dominant.
The planters’ success in lobbying for legislation which gave the landowner the priority lean
was crucial to their continuation of control (Hahn 1983; Wiener 1978). In the less fertile
regions where planters were less common and less powerful, merchants more frequently
became merchant-planters, which created a more revolutionary transformation of the social
structure (\Niener 1978:93). Although the study area is within a plantation area, it is the
later scenario, a merchant-planter transformation, that best describes the events at Vinton
during the postbellum period.

W. E. Trotter supplied farmers who owned their own land, tenants on other
persons’ lands, and tenants on his own lands. For many blacks, on their own for the first
time with little money, tools, supplies, credit, or meam to transport supplies the 10 miles
from West Point, the closest town, the Vinton store was probably their only means to begin
farming. Many of the other white landowners in this period became indebted to Trotter as
well. Trotter’s goods were shipped by rail to the county seat of West Point, and taken by
wagon to his store. He was supplied by companies from Memphis, Atlanta, New York,
Chicago, Mobile, and likely many others. He supplied a variety of goods, including
mules, to farmers on his own land. In 1886 he owned 32 mules, which he supplied for an
annual charge of from $45 to $100.

Other business at the Vinton store (all run by Trotter) included cotton ginning,
milling, and the Vinton Ferry. The Vinton Ferry had operated during the antebellum
period as well but in competition with the nearby Colbert and Barton ferries. In 1881
Trotter successfully petitioned for a change in the road network, so that traffic formerly
directed to the Barton Ferry was redirected to Vinton. Although the petition was opposed
by those who ran the Barton Ferry, Trotter’s request was granted. This change probably
increased Trotter’s revenues in two ways, first by the ferry tolls, and second, through

3 6
incidental store purchases made by ferry passengers. This arrangement continued until
1893, a year after Trotter’s financial collapse, when the Barton Ferry operators
successfully petitioned to direct traffic back to the Barton Ferry and discontinue the road
from the Barton Ferry to Vinton. It may be that 1893 thus nmrks the point when the Vinton

store ceased to be the major economic center for most Vinton residents.

Receipts from the Vinton store were sought, but few were located. Trotter’s
monetary strength, as assessed by R. G. Dun and Co., grew from the $5,000 to $10,000
range in the late 1860s to $10,000 to $20,000 in 1874, and to $20,000 to $40,000 in 1876.
In 1877 Trotter was the third largest merchant of the 52 merchants reported in Clay
County. By 1886 his business volume had increased to $40,000 to $75,000. Only 11
percent of all Southern merchants did this much business in 1880 (Ransom and Sutch
1977: 138). With this volume and his solid credit rating, Trotter probably could order large

quantities of goods on favorable terms from any national wholesaler.

Many of Trotter’s customers may have settled up their accounts periodically, selling
their cotton or other produce to him, or paying in cash. Unfortunately no individual
accounts from his store have been located, so it is difficult to compare cash and credit
prices, or to estimate interest rates. Trotter frequently required a chattel deed or mortgage
on the coming year’s crop before advancing goods to customers. These chattel deeds
specified the amount initially fumished, the additional amount Trotter agreed to advance
over the year, and the goods mortgaged in return. Usually these goods included livestock,
sometimes described in great detail, all the household goods owned by the family, and any
crops produwd by them or others for them. In some cases these deeds specified that the
lien applied to all crops, including subsistence crops and not just cotton or other cash

crops.

37

Chattel deeds to W. E. Trotter have been located in Clay County Chattel deed
books. These deeds show advances of $5,608 to 20 customers in 1877, $7,526 to 62
customers in 1878, $9,835 to 32 customers in 1879, $7,580 to 27 customers in 1880, and
$12,119 to 30 customers in 1881. Compared to the estimates of Trotter’s business volume
presented above, these chattel deeds must represent a small fraction of the store’s business
and certainly a small fraction of the total customers. The amount advanced to any one
farmer ranged from $1 to $492. The advances to blacks never exceeded $450, while J. J.
Cox, a white landowner who had many tenants on his own farm, was allowed to charge up
to $1,500.

Until 1882 these chattel deeds usually specified the landowner on whose farm the
crops were to be grown. By doing deed research, these farms have been located. Figure
2.4 shows the location, by section, of the number of farmers doing business with Trotter
from 1877 to 1881. Trotters main hinterland seems to have extended about 2.5 miles
north and south of the store, and about 4 miles west, for a total coverage of about 20
square miles. He may have also supplied some farmers on the east side of the Tombigbee,
especially since he kept the ferry in operation. However, no chattel deeds have been
located for Trotter in this area, which is in Lowndes County. Figure 2.4 should be
compared to Figure 2.5, which shows the number of black men age 1444, according to
the section in which they resided in 1870. This comparison suggests that Trotter fumished
a large portion of the bhck farmers in the general Vinton area, and did indeed have a
territorial monopoly (Ransom and Sutch 1977). Of the 70 chattel deeds located for

farmers included in this data set, 32, almost half, were to Trotter.

A number of families in the southern portions of Figures 2.3 and 2.4 would
probably have been fumished by the landowner J. J. Cox (mentioned above as one of

Trotter’s largest customers), as this area was known as Cox’s quarters. Individual tenants

38

 

 

 

 

 

 

 

IMila N
l 3 3 3
6
I
I0 3 4 4
7 ' Tr$er
ll House
4 T/6 3 3m |
TI?
ll
9
5 5
3 I

 

 

 

 

 

 

 

Figure 2.4. Number of Vinton store customers by location of farm.

3 9
may have come to the Vinton store and charged purchases to Cox’s account, or Cox could
have purchmed the supplies himself and redistributed them to his tenants. In any case, the
chattel deeds from Cox to Trotter suggest that Trotter also controlled, if only through Cox,
the fumishing of these tenants. Figures 2.4 and 2.5 may also be compared to Figure 2.3,
which shows Trotter’s land holdings. He certainly would have fumished all farmers on his
own land.

A collection of accounts from the Waverly plantation south of Vinton (Figure 2.2)
suggest that in the first few years after emancipation, most landlords bought supplies in
bulk to redistribute to their own laborers, many of whom may have been true sharecroppers
working for a share of the crop as a wage. A number of Clay County planters registered
contracts with the Bureau in 1865, including Jesse Dukeminier with 16 hands and 19
dependents, Joseph Fields with 17 hands and 13 dependents, F. E. Harris with 15 hands
and 13 dependents on the Cox quarters south of the Vinton store, and Sherod Keaton of
Vinton with 15 hands and 12 dependents. Elisha Strong, a large planter during the
antebellum period, registered contracts with up to 75 hands. These contracts would have
been for some form of wage labor. A contract from the Sharpley’s Bottom plantation in
neighboring Monroe County (Figure 2.2) specified that work would be conducted in
squads, with 15 acres per person cultivated, works days of from 9 to 11 hours depending
on the season, and fines for days not worked. The wage was one-fourth of the crop (Kern

et al. 1982).

By the 1880s, it was more common for individual families to establish independent
relations with a merchant, although the landlord might be needed to endorse the charges
(Adams 1980: 105). This corresponds to the increasing autonomy granted tenants over the
postbellum period, and to the increasing popularity of share tenancy and renting, over
sharecropping and straight work. As in most areas of the South, the lack of cash,

4,0

 

 

 

 

 

 

lMiIe
‘——"“ 5 6
~15
L—____.‘. 4
“ '5) 776
TI?
15 IS 4
+20/ 1
3; E

 

 

 

 

 

Figure 2.5. Number of adult black males by location of farm, 1870.

41

financial uncertainty, and freedmen’s unfamiliarity and dissatisfaction with wages led to a
fairly rapid progression into other systems. The squad system mentioned above, in which
a group of persons worked one area of land together under the general supervision of one
of the group, can be documented in the study area at the Thomas Martin plantation near
Waverly, from 1869 until 1872 (Adams 1980; Clay County Case File 22). However,
farming under this system at the Martin plantation proved entirely unsuccessful. After
several years of net losses, the plantation was reorganized, and in 1874 contracts were
made with individual tenants who farmed on a household basis. Cotton production, which
had reached a low of 31 bales on the 2,826 acre plantation in 1872, increased to 139 bales
by 1879. Estate reports also suggest that sometime between 1876 and 1879, tenants on
the Martin plantation may have begun to farm on a renting rather than sharecropping basis,
although rent was still paid from the sale of the crop (Clay County Case File 22).

The speed at which household-based tenant farming, versus wage labor or the
squad system, was adopted within the study area is not known. There was probably a
good deal of variation within the region. The Martin plantation seems to have switched by
1874, but the squad system was still in use on the B. A. Duncan plantation, on the western
edge of the study area, in 1875 (Clay County Case File 236). Full standing rental
arrangements were in place at the Sharpley’s Bottom plantation quite early, by 1868, where
households rented 30 to 50 acres parcels for a set amount of cotton. Interestingly, some of
the first rental contracts at Sharpley's Bottom involved cases of multiple households
farming together. These had split up into individual household farms by 1870. The
landlord referred to the tenants’ housing as the ”quarter ”in 1868, but referred to their
”houses” in 1870. The tenants at Sharpley‘s Bottom had free use of a large section of cane

and other unimproved land, which was important in raising livestock (Kern et al. 1982).

42

It is certain that some form of renting was quite common by the mid 1870s, when
chattel deeds become quite numerous in Clay County. These deeds represent a lien given
by the tenant on the crops they are to grow. The household-based system was undoubtedly
firmly established by 1880, the initial time period used for the analysis on household
composition and farming.

The chattel deeds rarely mention a house as security, suggesting that most of the
tenants had their house provided for them along with the plot of land, as was frequently the
case. Records from the Martin plantation mentioned above also support this interpretation,
as entries in 1868, 1874, and 1876 concern building or repairing tenant houses.
Ownership of the house was an additional lever landlords had over tenants. One Vinton
tenant farmer took a landlord to court, in part because he could not pick all of his cotton
crop because the landlord had turned him out of his house in January. His landlord
claimed that the tenant should have completed the picking in December, but the tenant
testified that most hands on the plantation had not finished picking their cotton by this time,
due to an unusually rainy season (Clay County Case 236).

Little specific information is available about details of tenant contracts and
arrangements within Vinton. A court case in the 18805 provides some information about a
tenant living just north of the Vinton store. This tenant, Henry Keaton, had been a slave of
Sherod Keaton, a prominent Vinton resident. Upon emancipation, Henry remained on the
Keaton farm and rented about 25 acres from the Keaton family. He also had use of about
10 acres of sedge or marsh land just south of his farm, at no charge. In 1879-1880 he
rented by the share method, with his rent being one-third of his corn and fodder and one-
fourth of his cotton. This was a common system in the Vinton area as in other parts of
Mississippi, often called renting on thirds and fourths (Adams et al. 1981; McClurken and
Anderson 1981; Smith, Barton, and Riordan 1982). It was probably restricted to those

4 3
tenants who furnished a good deal of their own supplies, animal power, tools, and other
necessities, with the landlord taking a larger share when the tenant had to be more
thoroughly supplied. The produce that went to pay Henry‘s rent was 50 bushels of com,
for which he received $0.45 per bushel; 300 pounds of fodder for which he received $1 per
hundred; and 1,180 pounds of cotton worth $0.1175 per pound. This would have yielded
a total rent of$l64.15 for that year.

There were two cabins on Henry‘s 25 acres, one 15 feet by 16 feet and one 16 feet
by 17 feet, both 8 feet high and made of wooden planks. One of these Henry built, for a
cost of $ 14. The other presumably was built by an earlier tenant or the landlord. Henry
also bought a smokehouse from a neighbor and had it moved onto the property, and with
some other farmers he built two cotton storage sheds in the middle of his fields. There
was also a com crib and a stable on the property. The Keaton household at this time
consisted of Henry, age 60 and widowed, a son age 16, and 3 daughters, age 15, 10, and
8. The agricultural census listed 20 acres improved: 10 in cotton, from which they
produced 4 bales; and 10 in corn, from which they produced 100 bushels. The 50 percent
of this acreage donated to cotton production is relatively low for the region (see below in
chapter 5). The estimate of the 100 com bushels may also have been low. Henry’s one-
third of corn rent that year was 50 bushels, suggesting a yield of 150 bushels. The court
case also suggested that Henry rented another 5 acres in addition to the twenty mentioned
above. Part of this other five acres may have been used for the animals and planted in
vegetables, especially since the census lists his household as producing 25 bushels of
sweet potatoes. The household owned 1 mule, 4 cows, 25 pigs or piglets, and 30

chickens.

Henry supplemented his income by helping repair fences and outbuildings on the
Keaton lands. In exchange, he was allowed to gin his cotton at the Keaton gin, free of

44

charge. Ginning charges were usually about $4 a bale (Henry produced 4 bales in 1880,
according to the agricultural census). Sometimes he was paid for additional labor, such as
in 1870 when he helped with miscellaneous repairs. For this work he was paid $0.75 a
day for splitting rails, $1 a day for putting up fences, and $2.50 a day for hauling rails.
This $2.50 included the horse or mule he provided.

Henry Keaton’s landlord was Mary Cogdell Keaton Richardson. Her brother
Thomas Cogdell helped her run the farm. A Vinton column in the West Point paper quoted
a relative, in the following manner:

Our younger brother, J. M. Cogdell, says he can make cotton as cheap as any
man as long as he can keep his smokehouse and corn crib at home. He killed
five pigs last week, netting him 1,200 pounds. He claims not to have bought

any meat in six years. This rs why he can raise such cheap cotton [cotton
prices that year were expected to be around $.04 or $. 03 a pound] (Wat

W, 18 February 1898).

The fact that the Henry Keaton household owned a number of cows and pigs
suggests they may not have had to buy much meat. Henry Keaton appears in the chattel
deeds to W. E. Trotter, whose store was just south of his farm, for the low amount of $90
in 1879 (the agricultural year that the 1880 census come closest to measuring). However,
many other tenant farmers did not own as much livestock as the Keatons and probably ran
up much higher bills at the store. The $90 figure is low when compared to the mean
amount advanced that year, $307. The com to cotton ratio for the 21 black Trotter
customers who could be located in the 1880 agricultural census was 1 to 2.64. As was

noted above, Henry Keaton and his family planted a 1 to 1 ratio of corn and cotton.

It is not known if W. E. Trotter specified the acreage his tenants were to plant in
cotton, although this was a common practice for landlords, especially when tenants farmed

on the sharecropping methods, and merchants, in some case with the explicit purpose of

4 5
restricting self-sufficiency(Fer1eger 1984: Fite 1984; Nixon 1938; Ransom and Sutch
1977). Fite (1986) has suggested that most share tenants (which would characterize
Henry Keaton) were able to produce a good portion of their food, and that it was
predomimntly sharecroppers who spent up to half of their income on food. Records from
the Sharpley’s Bottom plantation north of Vinton show that while most farmers were
tenants, their level of livestock and tenant accounts suggest a good deal of self-sufficiency.
However, under a new landlord who enforced a preference for sharecropping, increased
cotton acreage, and did away with use of unimproved lands, 9 out of 12 households
exhausted their yearly store of com by April, with 2 more running out in July (Kern et al.
1982).

At Vinton, Henry Keaton may have enjoyed an unusual freedom to plant less cotton
than most tenants. His landlord, Mary Cogdell Keaton Richardson, and her brother
Thomas, who helped run the farm, may have shared the sentiments of their relative J. M
Cogdell, u indicated above, and allowed if not encouraged diversification and self-
sufliciency among their tenants. The Henry Keaton households’ charges to the Trotter
store were even lower in surrounding years; $38 in 1877, $24 in 1880, and $25 in 1881.

It should be noted that the 4 bales of cotton that the Keaton household produced on 10
acres was the highest yield per acre of any of Trotter’s customers for that year, and that the
100 bushels of corn on 10 acres was also a relatively high yield per acre.

Other farmers recognized the importance of diversification in an economy in which
cotton production was the mainstay. Production of corn, other crops, and livestock had all
decreased during the postbellum period while the production of cotton increased, both in
typical plantation areas and in non-plantation areas and among white as well as black
farmers (Ferleger 1984; Fite 1984; Hahn 1983; Weiman 1985; Wiener 1978; Wright 1985).

However, diversification was often encouraged. Many local residents belonged to the

46

Grange, an agricultural organimtion which discouraged over reliance on cotton. The
Grange publication Wm whose goal was ”to improve the Agriculturalist
in his moral, social, intellectual, and pecuniary condition,” was published in nearby
Columbus, Mississippi. It railed against the Mississippi Valley Cotton Planter’s
Association and the furnishing merchants, and the demands of northern capitalists for more
cotton. Lecture circuits were a common strategy of the Grange, and a Vinton resident, W.
H. Dukeminier, was a Deputy Lecturer.

It is certain that many tenants did not enjoy Henry Keaton’s level of agricultural
wealth, or stability. Studies of tenant mobility suggest that in some area, as many as one
half of all tenant farmers relocated each year (Cooper 1933; Johnson 1934; Mandle 1978;
Rosengarten 1974). Account books and court depositions relating to the Sharpley’s
Bottom community provide information on several tenants. One, Jerry Harris, moved 6
times between 1986 and 1901 , although all moves were within a two or three mile radius.
Another tenant moved seven times from the early 18803 to the late 18908 (Kern et al.
1982). In contrast, Henry Keaton seems to have been along term resident on the Sherrod

Keaton property in Vinton, where in fact he had been a slave.

A very different pattern is seen in the relationship between the tenant Sam Smith
and his landlord R. B. Duncan, a large landowner in the western portion of the study area
in 1875. Smith’s contract, which was actually negotiated by B. A. Duncan, as agent for
R. B. Duncan, stated that Smith was to receive one third of the cotton, com, and fodder he
produced during 1875, with the landlord receiving two thirds. The landlord agreed to
fumish Smith and his squad of two persons 180 pounds of meat free of charge. Smith
was to provide ”good and faithful labor” in seeing to his crops. The contract stated that if

Duncan felt Smith was not giving proper attention to his crops, Duncan could hire extra

47

labor and charge it to Smith’s account. The contract further stated that Duncan had a lien
on and control of the crop until he and Smith had settled their contract.

Contracts such as this one are rarely preserved, and in fact this contract was located
only because it had been given as evidence in a court case. Smith’s testimony suggested
that although he had signed the contract with his mark, he did not understand it. For
example, when his three bales of cotton were taken to the gin in West Point, he
immediately turned over the one bale he felt he was owed to a local merchant, as payment
on the merchant’s lien on Smith’s wife’s land. However, Duncan felt that Smith had been
negligent toward the crop and owed Duncan some payment for lost cotton, meaning that
their account was not firlly settled and Smith did not yet have a right to his share of the
crop. Duncan thus intervened and denied Smith’s right to turn over the one bale of cotton

to die merchant.

Smith also showed that he did not fully understand the furnishing agreement within
the contract. When asked during the trial about the terms of the contract, he stated that the
landlord ”was to give me one third of all I made and feed me free of charge.” Smith
actually used less meat than the 180 pounds allowed him in the contract, and for which
Duncan later agreed to credit him. But, he was charged for meal, molasses, and other food
items, which may have come as a surprise to him. Smith’s lawyer also pointed out that
Dmcan charged Smith $0.10 per pound for flour, when it was generally selling for $0.05
to $0.06 per pound.

The series of misunderstanding between Smith and his landlord, and the merchant
who had originally held the lien on his wife’s land, resulted in Smith not being able to clear
the lien. In frustration at his landlord’s denial of his right to the one bale of cotton, Smith

left the area and went to work for several months in the western part of the state, probably

48

as a wage laborer on a large Delta cotton plantation. Smith testified that before he left, he
understood the merchant as saying Smith could pay on the lien whenever he saved some
money. Smith claimed to also have written the merchant several weeks after he had left
town, to tell him of his intentions and progress. It is possible that having been evicted
from the Duncan plantation, so that he could not gather the remainder of his cotton (which
may have already been ruined by wet weather), and having been warned of the charge of
neglect against him by the landlord, Smith may have felt tint going to work in another area
was his best chance of raising the needed money. One would like to read the letter Smith
wrote to the merchant, although it is not included in the court papers. It appears,
however, that the merchant did not take kindly to Smith’s absence, for he sold the lien he
held on Smith’s wife’s land to Smith’s old landlord, the Duncans. The final ruling in the
case suggests that after several years, the Duncans continued to hold the lien. However, it
is not clear that Smith ever managed to recover any payment for the season’s work at the
Duncan plantation.

This case is an example of a sharecropper who was completely fumished by his
landlord, bringing only labor and being paid a share of the crop as wages. Although some
tenants undoubtedly continued to farm in this manner, and be completely fumished by
plantation owners on a large plantation, it is likely that increasing numbers of tenant
farmers farmed in a slightly more autonomous manner, under the household-based system

in which they made their own agreements with the local Vinton store.

The Vinton store continued as the major fumishing center in Vinton until the early
18908, when a series of suits against W. E. Trotter’s son W. T. Trotter, eventually also
ruined W. E Trotter. W. E. Trotter ran the Vinton store mostly with the help of his sons,
who continued in the family occupation as merchants. His son James married the daughter

of an Aberdeen merchant (see Figure 2.2), and set up a store in the crossroads community

49

of Muldon, Mississippi, between Aberdeen and Vinton. In 1883 another son, W. T.
Trotter, age 29, was engaged as a clerk in a West Point store. He probably began his own
store in West Point around 1884. By 1888 another of W. E.’s sons, Ben, was working in
W. T.’s West Point store. W. E. Trotter also did business with this daughter’s husband, a
Mr. Kirk.

Son W. T. Trotter opened his store in West Point with about $4,500. Although his
stock had increased to $7,000 by 1886, he was described as financially ”cramped” in 1887.
At the same time, W. E. mortgaged several of his properties, perhaps in an effort to funnel
money to his son’s failing business (see Figure 2.3). From 1888 to 1892 a number of
suits were brought against W. T. Trotter. W. T. was also charged with ordering goods
using his father’s superior credit rating. More damaging were charges that W. T. had
attempted to avoid payment for goods ordered and received for this West Point store. He
was accused of conspiring with his father W. E to transport the goods secretly to the
Vinton store where they were intermingled with goods W. E. Trotter had ordered and paid
for. The charge was that the Trotters hoped to confuse W. T.’s creditors and deny receipt
of more tlmn $10,000 worth of goods.

W. E. Trotter was initially cleared of any involvement, but an appeal to the
Mississippi Supreme Court brought the ruling, on 13 May 1892, that ”the circumstances
tending to show confederacy are too strong and too numerous and those supporting his
good faith too few and inconclusive to warrant a decree in his favor.” W. E.’s Vinton
property thus became liable for the son’s debts and was sold at auction in September,
1892, by which time the family had temporarily fled to Louisville, Mississipi, about 45
miles away (Figure 2.2). By 1896 they had returned to Clay County, and W. E. died in
Vinton in 1899 at the home of his daughter Fannie Kirk.

5 0

Testimony in these court cases provide some information about the Vinton store
itself. For example, it contained one Lille safe, three showcases, six counters, one piano,
and two writing desks. The property also included one steam boiler and engine, one 60-
saw (blades or teeth) gin, with feeder and belt, and one Straub gristrnill and belt. This
property, plus the land in Section 36, was purchased by W. Porter Rankin, owner of the
Rankin Manufacturing Company, maker of pants, overalls, and so forth of Nashville,
Tennessee. It is likely that Rankin had been one of Trotter’s suppliers. Rankin visited the
area periodically, to check on business and hunt with local residents, but be either
employed someone to run the store or leased it out from 1892 to 1900. In 1897- 1899 it
was run by a George Neville from Jackson, Tennessee, probably an associate of the
Rankin Company. The Vinton store now became a subsidiary of Rankin Manufacturing
Company. It was listed in 1899 and 1900 by the R. G. Dun and Company agents as doing
business in the $20,000 to $40,000 range.

The breakup of the Trotter holdings had important ramifications. W. E. Trotter had
been a central figure in the Vinton community. He was the person most frequently called
on to handle business affairs, serve as a witness, help settle estates, and perform
neighborly services. His vast acreage was purchased by several parties. Although
Trotter’s policies at the Vinton store are not documented, it is likely that passage of the
store to an absentee owner meant changes in the extension of credit and the way business

was conducted.

Rankin kept the Vinton property until January 1900, when he sold it to a local
landowner, Henry D. Watson, I. The Watson family was from Monroe County, just north
of Vinton, but had previously owned some property in the area during the antebellum
period. Henry D. Watson lad bought part of the Cox quarters south of the Vinton store in
1894, and also later acquired extensive acreage in the western part of Vinton and the

5 1
community of Strongs, Mississippi, several miles west of Vinton. Deed research suggests
that the Watsons owned somewhere around 2,000 acres, a sizable holding although not
half of what Trotter had controlled. Members of the Watson family lived in the old Trotter
house at Vinton until 1941. They continued to operate the Vinton Store for many years,
although it now became more of a commissary for farmers who rented land from the
Watsons under the standing rent arrangement. Account books for 1901 and 191 1 show no
accounts for independent landowners. The Vinton store was no longer listed in the R. G.
Dun and Company ledgers, which suggest that the Watsons obtained more of their goods

from local sources compared to Trotter, and did a smaller volume of business.

Two account books of the Watson family, one dating to 1901 and one dating to
191 1-1912, suggest that like Trotter, the Watsons supplied a variety of items in the Vinton
store, including food, dry goods, agricultural supplies, clothing, medicines, and
miscellaneous items (Watson Collection n.d.). They also took care of bills to other persons
or institutions, such as fines, legal fees, donations to a local minister, loans to friends or
family members, payments to others for labor on the crop, bills to another merchant, or
bills to the doctor, or in one case, a final charge for a coflin. The account kept for the
Watsons themselves suggest that they raised a considerable amount of hogs. Whether
these were all for home use or sold to the tenants is not known. The store also carried
other meat in addition to the staple of pork; beefsteaks and roasts were noted in the house
account for 1912. The Watsons purchased corn from some of their tenants, and hired them
for temporary labor. They often supplied cash to tenants, which was charged to the
account, and sometimes noted charges that the tenant had agreed to ”work out,” some
charge. The books show an entry in almost every account for ”cash for Christmas.” This
custom has often been described as a bonus that tenants came to demand. While a cash
advance was made available to the tenants at this time, it is perhaps misleading to

clmracterize it as a bonus when it was charged to their account.

52

Comparison of prices is diflicult because many entries did not specify the quantity,
but might only list ”meat and meal.” However, comparison of the minority of cases that
did specify unit prime suggests that the tenants paid an extra charge for their credit, a was
customary. For example, the standard charge for meal in tenant accounts in 1911-1912
was 75 cents per bushel, while Henry D. Watson charged himself between 55 cents and 66
cents per bushel.

Watson seemed to have used the middle of December to end one year and start the
clnrges for the next. At this time the charges were tallied, as were credits. The credits
were mainly from cotton turned over to the store, but some accounts would also have other
credits, such as from labor supplied during the year. Cotton was separated by quality,
with the higher qrality credited at 10 cents a pound and the lower quality at 8 and three-
quarters a pound. The total monetary value of the cotton was then subject to a processing
charge, which consisted of charges for shipping, weighing, storage, and insurance. The
total processing charge was usually about six to eight percent of the value of the cotton.
The resulting cotton credit, and whatever other credits the tenant might have, were then
applied to the tenant’s bill. The bill was composed of all store charges, plus 20 percent

interest on these charges, and the yearly rent. These accounts are studies in more detail in
Chapter 5.

If 191 1- 1912 was a representative year, most of the Watson’s tenants became
trapped in the system of debt peonage that has generally been used to describe the
fumishing system. However, the Watsons allowed some tenants, and other blacks, to
purchase land. One account in the 191 1- 1912 ledger included a note that a Pearce Mealer,
one of their customers, had paid $43.80 as part payment on 80 acres acres of land. The
balance due on the land was $456.20, indicating a total price of $500. Earlier, in 1897,

53

the Watsons sold Larry Keaton 40 acres on the Cox's quarters section of their property.
Larry was a son of Henry Keaton and a member of the 1880 Henry Keaton household
discussed above. He paid for this property with a down-payment of four 500.pound bales
of cotton, with 12 more bales to be paid over the next three years, a common arrangement.
The average cotton price from 1897 to 1900 was 7.14 cents per pound (U.S. Census
Bureau 1960). Multiplying this by 16 bales of 500 pounds each gives a purchase price of
$57 1.20.

These purchases were part of a slow increase in black farm ownership around the
turn of the twentieth century. This trend is also documented within the study area in
Chapter 5 below. Most of these holdings were 40 to 60 acres, financed by multi-year
mortgages. Deed abstracts were searched for the names of other Watson tenants or close
relatives. This task suggested that two other tenants who had farmed with the Watson
eventually acquired land. These persons were Selvin and Andrew Lenoir. Figure 2.6
shows these and other new landowners around the turn of the twentieth century. Except
for C. E. Benton, the Henry Wilson family, and J. F. Ellis, most of these landowners are
black farmers who purchased small acreages.

These purchases are in a sense extraordinary. Most black tenant farmers could
barely come out even at the end of a year, let along save enough to buy property. The
circumstances of most of these families are not known. Larry Keaton’s family background
has been discussed above. Andrew Lenoir, one of those shown in Figure 2.6, was a local
schoolteacher, and so would have earned a small cash salary. A study of several cotton
producing counties in Texas just after World War I found that 63 percent of all black
landowners had depended on assistance by family members to purchase their property
(Sanders 1922). The stability of these purchases in not known. Several suits involving

payment from some of those farmers shown on Figure 2.6 suggest that not all of these

54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Levi Holland
C.E. Benton Jcbe Trotter
k
-2
0
Henry *
Wilson 5
Fomtty -'
a.
25 25 8’
so“
Srntth Bryant thlte
Matthews Mitchell Guttin "3""
Chas. Thame
Geode _ Doc
C7101 Willi. '
Jonnson 0"” S rmq Henderson .
H. Brooks . Henry
RICO Grittin 35 Rice 36 “I." 31
Jettenon
Dauqhe r ty J F- Mace
Ellis MIJCQ‘ ,
Ander - Henderson
son the
Hodge Henderson Lory ‘
the m f
0' f.
Selvrn ..
:9 L“! O. Lord
.. . .0 “'0‘ in
AD
Leno"
2 l 6

 

 

 

 

 

Figure 2.6. New landowners in study area.

55

black landowners were able to retain their land for long. This is consistent with national
trends. The Census Bureau noted that one of the strongest trends in farm ownership from
1910 to 1930 was the loss of acreage by negro owners (U.S. Census 1933).

These small increases in property ownership at this time were also accompanied by
increases in personal property. Most blacks had very few material possessions upon
emancipation, and landlords fumished most household goods. Two Clay County personal
property tax rolls have been located, one for 1886 and one for 1902. Figure 2.7 shows the
distribution of personal property for District One of Clay County, which included most of
the Vinton community. A small portion extended north into Monroe County. This figure
shows that from 1886 to 1902, the percentage of persons with either no taxable property or
property worth over $ 1000 declined, while the percentage of the population having taxable
property valued at $101 to $1000 increased. While these taxes are generally for luxury
items like clocks, watches, vehicles, weapons, silver plate, pianos, and similar items, they
probably provide some indication of a slight increase in general household wealth over the

period.

56

.358 $6 ._ 8.7590 .3385 28088 no 8339me

ooomloo; coo—low» cowl—o; 00...;

  

 

 

.-

 

‘-

 

 

 

 

 

  
 

woo. M

em? I

  

5m 8%:

O

 

 

 

 

«

    

To

1..- om

 

CHAPTER 3
HOUSEHOLD COMPOSITION AND LIFE CYCLE

Intr;o__ductign

This chapter presents the analysis of household composition and life cycle.
Households, as basic units of kinship, residence, and production, are analyzed according
to composition and structure. However, since households change over time, in
accordance with events like marriage and childbearing, the household life cycle approach is
also utilized. And finally, households are looked at from the perspective of individuals

within the household.

At the outset it may be useful to define several terms and to look briefly at the age
structure of the population. Instead of fitting households in the study area into a standard
typology, which generally tries to encompass several variables at once, a basically
descriptive system has been developed. This system describes households according to
five variables -- their core type, their stage (maturity of household), the presence of lineal
relatives, the presence of lateral relatives, and the presence of non-relatives. The core type
of household is defined as one of the following: 1) a married couple; 2) a married couple
with children; 3) a single parent with children; 4) a group of adult siblings; 5) a group of
other kin; 6) a group of non-kin; or 7) persons living alone (solitary households). Second
and later marriages are referred to as serial marriage. Households having a married couple
as the core are called ”complex” households when one of the spouses is serially married,

i.e. has been married before.

37

CHAPTER 3
HOUSEHOLD COMPOSITION AND LIFE CYCLE

laments;

This chapter presents the analysis of household composition and life cycle.
Households, as basic units of kinship, residence, and production, are analyzed according
to composition and structure. However, since households change over time, in
accordance with events like marriage and childbearing, the household life cycle approach is
also utilized. And finally, households are looked at from the perspective of individuals

within the household.

At the outset it may be useful to define several terms and to look briefly at the age
structure of the population. Instead of fitting households in the study area into a standard
typology, which generally tries to encompass several variables at once, a basically
descriptive system has been developed. This system describes households according to
five variables -- their core type, their stage (maturity of household), the presence of lineal
relatives, the presence of lateral relatives, and the presence of non-relatives. The core type
of household is defined as one of the following: 1) a married couple; 2) a married couple
with children; 3) a single parent with children; 4) a group of adult siblings; 5) a group of
other kin; 6) a group of non-kin; or 7) persons living alone (solitary households). Second
and later marriages are referred to as serial marriage. Households having a married couple
as the core are called ”complex” households when one of the spouses is serially married,

i.e. has been married before.

57

58

This core household can then be modified by the presence of linear relatives, in
which case it is extended. If the household contains lateral relatives, it is expanded. If the
household contains non-relatives, it is augmented. Each type of addition to households
was given a series of codes depending upon the type of person or persons added to the
household. For example, the extension category was coded 1 for no extension, 2 for
mother, 3 for father, 4 for both parents, and so forth. The complete list of extension,
expansion, and augmentation categories is given in the Appendix. The categories of
extension, expansion, and augmentation are not mutually exclusive, and each household is
coded separately for each type of addition. The only typologies] terms used in describing
the households are ”nuclear” family household, meaning a married couple with or without
children, and the ”stem household” composed of a household core of parent(s) extended by
a married child and spouse.

Households are also described by stages of maturity. These stages are based upon
a combination of the age of the wife (or head if female) and the ages of children in the
household, and are designed to measure the household life cycle. The mother’s age is
used instead of the father’s since it is more closely tied to the physical reproduction of the
household (mother’s age was also used by Shifllett [1975] although the present analysis
did not follow his stages). Household stage is based entirely on the core household
members, and is independent of extension, expansion, or augmentation. Although these
household stages do not actually follow any one household through time, they help
overcome the static nature of typological analysis. The complete breakdown of stages,
either 1 to 5 or in some cases 1 to 3 to reduce the number of potentially zero cells, is given
in the Appendix. In all cases Stage 1 is the youngest stage, and either Stage 3 or Stage 5 is
the oldest stage.

5 9

In addition to these classifications, households are also described by the number of
persons of various categories within them. These categories of persons are nuclear family
members, non-nuclear family members (called ”extra” family for convenience), non-family
members, and total ”extra” members, which is the sum of the extra family and non- family
members. Household members are also classified as consumers or workers, for analysis
of dependency ratios. Many systems of classifying consumers and workers exist. This
analysis utilizes the same parameters as used by Chibnik (1986). This scale ranks males
and females on a scale of 0.1 to 1.0 by levels of consumption and production, according to

age. It is presented in the Appendix.

Age structure is a basic aspect of a population that can have many long-reaching
ramifications. Figures 3.1 thru 3.4 illustrate the age structure of the Vinton and West Point
samples. Since the sample sizes vary considerable from subsample to subsample, each
subsample is plotted separately. This allows easier comparison of the general age structure
of each population, regardless of differences in Ns. Each horizontal bar represents a five
year age interval. The first bar represents the 0-4 year olds, the second bar the 5-9 year
olds, and third bar the 10- 14 year olds, and so forth until the final bar, which groups all
persons age 70 and older. There is considerable irregularity in the pyramids, which is
partly the result of small sample sizes. Yet, these figures point to important differences
between the black and white samples, and between the urban and rural samples. One of
the more ovbious differences is the relatively high number of persons in the 0-4 and 5-9
age groups (the bottom two horizontal bars) for the Vinton population. This is congruent
with the high fertility of this group, as discussed in Chapter 4.

Another striking finding is the relatively low number of middle aged black men,
both in the Vinton and West Point samples. This is true when the age structure of black

men is compared to white men or to black women. The reason for this is not known,

60

1880 blacks. Vinton

—“
l l l J 1 1
F I ' 1 I V I

 

-400 -300 ~200 - 100 0 100 200 300 400 500
males females

1910 blacks, Vinton

 

-
-
—
-
_
_
_
_
—
_
—
_
_
—
—
-300 -200 -100 0 100 200 300
males females

Figure 3.1. Population structure, Vinton, black sample.

61

1880 whites, Vinton

%
V

-60 -50 -4O -30 -2O - 10 O 10 2O 3O 40 50
males females

I J .l l A I A l L L
r I“ U I V V V I V V

I
fl

1910 whites, Vinton

l 1 l
v v 1

-4o -30 -2o -10 o .10 20 30 40

males females

Figure 3.2. Population structure, Vinton, white sample.

62

1880 blacks, West Point

 

males females

1910 blacks, West Point

I J l A l
V V V V V

j A
Y Y

j
V

J.

-80 -60 -40 -20 0 20 4O 6O 80 100

males . females

Figure 3.3. Population structure, West Point, black sample.

63

1880 whites, West Point

 

L L 1 l 1 J
r v 1 v v 1

~20 ~15 ~10 -5 O S 10 15 20
males females

1910 whites, West Point

 

_
—
_
—
_
_
_
—
_'
_
_
—
—
p . _
~80 ~60 ~40 ~20 0 20 4O 60 80
males females

Figure 3.4. Population structure, West Point, white sample.

64

although the Civil War is a possible contributing factor, both from mortality and from
higher out-migration by black men, compared to black women. This effect would be more
likely for the 1880 sample. Another obvious feature of the age structure is that there are
more middle-aged black women in West Point than in Vinton. This is probably because of
the ease with which black women could find jobs in towns. However, for most age

groups there are still as many or more black women than black men in Vinton.

The white population generally shows a fuller age structure, indicating relatively
more persons in the middle ages. There is a reduction of males in the 35-39 and 40-44
group for the 1880 Vinton sample, and in the 40-44 and 35-39, 45-49, and 50-54 groups
for the 1880 West Point sample. These age groups would have been 15 to 39 years old
during the Civil War, and their reduction in 1880 may be the result of war-related mortality
or morbidity.

One of the important findings of this exercise is that there is not a great deal of
change in the age structure of any group illustrated from 1880 to 1910. Generally there is
more variation between the different subsamples for 1880 or for 1910 than between one
subsample over time. This is important to keep in mind when comparing household
compositiornl changes over time, and in looking at changes in fertility from 1880 to 1910.
It is suggested that change in age structure is probably not a major explanatory variable for
this population. A consistent change in the age structure between the ages of 20 and 50
would be of most importance in comparisons of fertility. There is a noticeable drop in the
04 age group from 1880 to 1910 for several subsamples, which is congruent with the
decline in fertility documented in Chapter 4.

65
H l 119 '|'

Tables 3.1. and 3.2. summarize the numerical descriptions of household
composition. The variable total persons is more homogeneous within each time period
than within the various racial/tenure groupings. For example, one dominant trend is the
reduction in household size over time, for every subsample. The smallest mean household
size in 1880, 4.7 persons for white sharecroppers, is near the largest mean household size
in 1910, 4.75 for white renters. It can also be seen in Tables 3.1 and 3.2 that the largest
reduction in household size is contributed by the categories extra family and non- family,
which combine to make the category extra persons. Extra persons decline from a range of
1.7 to 2.6 persons in 1880 to a range of 0.33 to 0.53 persons in 1910. This decline is so
severe that it has a larger effect overall than the change in nuclear family size, which
actually increases over time. These trends are also illustrated in Figures 3.5 and 3.6.
These figures illustrate the extreme declines in extra family and non-family persons in
households, as well as the overall decline in total members and increase in nuclear family

members.

The numbers of consumers and workers were calculated only for the 1880 sample,
so they could be used in conjunction with the analysis of agricultural production. These
measures show slight variation by racial and tenure subsamples, with less variation in the
summary measure of the dependency ratio. A final variable presented in Tables 3.1 and
3.2 is mean age of household head. Owners are, on the average, older than renters, for

both blacks and whites. This pattern is stronger in 1910 than in 1880.

Tables 3.3 and 3.4 present information on type of household core. It is
immediately apparent that the nuclear family household (types 1 or 2) prevailed. At least 70
percent and usually higher percentages of all tenure and racial groups were formed around a

66

 

 

w..—
2.—
cm.—
hm.—
hm.—
mm.—

33
can...

wa

nod

. and

ooN

~w.~

KEN

nmd

and

$6

Med

36

3.6.

no.
on.
_._
co.
3.—
_._

has.
-8:

N..—

9m

QN

9N

Wm

ed

meow—en
«.58

2

~.—

m.—

me.

me.—

n.—

v.—

ease
E0
2

o...
3
3.
S.
3.
3.

base

.52
z

a.”
9m
3
an
3
3

base

.828:
Z

54..

c6

9n

ad

ad

«6

38.8.—

.58
Z

mdm
ode
9.3
adv
nae
ode

3
5:0 one

alas

M3 53

 

.23.: we cognac .8». .8582: 20:399.. mucous—30$: _.n 035—.

 

67

.c3c_> .2: xva .o_o_uom®_ .coEanEou cc no comtmanu .m.m 0.59“.

 

XE zoz

 

99.me E
89.35 D
0.9.305 I
82.305 I

 

 

 

 

Eocwmsoc E comaoa .8 5098.8
2.; <56 5582 452

 

 

 

68

.c3c_> 6:: 3E; .o_o_uow®_ .coEBanoo cc no comCmQEoU oh 8:9“.

 

Eocmmsoc E comama no 5388
xi zoz Z<.._ «$50 «(@4032

 

053%; E
083%; D
99.3013 I

owe _ .3013 I

 

 

 

itch

 

 

 

 

69

Table 3.2. Miscellaneous household measures, expressed as means, 1910.

 

Group ageofhh N,total N, nuclear N, total N, extra N,extra N, non-

 

 

led persons family family family persons family
919.! 48.4 4.52 4.07 4.15 .37 .45 .08
m 39.4 4.34 4.01 4.26 .08 .33 .25
m 47.66 4.31 3.90 4.21 .09 .41 .31
m 40.9 4.75 4.22 4.46 .29 .53 .24

 

 

Table 3.3. Household types, by race and tenure groups, 1880.

‘ (table gives percentage of each group in following household types)

 

GmupN typeltypeZtypetype3 typetype4 typeS W6
nuclear mrclear 1+2 single 1+2+3 adultsibs other-kin nonkin solitary

type7

 

 

 

 

 

+kid par
1219122 4.5 81.8 86.3 9.1 95.4 0 O 4.5 0
M 12.8 72.3 85.1 9.0 94.1 .5 .5 1.6 3.2
M 12.5 70.8 83.3 11.5 94.8 .5 .5 2.1 2.1
mg 3.1 67.2 70.3 14.1 84.4 7.8 1.6 1.6 4.7
M 3.8 69.2 73.0 11.5 84.5 3.8 0 3.8 7.7
£213.12 16.7 66.7 83.4 o 83.4 o o 16.7 0
Table 3.4. Household types, by race and tenure groups, 1910.
M of households from each race/tenuflroup in following household types)
Group N type 1 type 2 type type 3 type type 4 type 5 type 6 type 7
msclear nuclear 1+2 single 1+2+3 adult other non solitary
+kid parent siblings kin kin
M 14.9 64.2 77.1 3.0 80.1 1.5 3.0 3.0 10.4
W 18.7 62.8 81.5 11.0 92.5 1.0 1.9 .2 4.4
M 8.9 66.7 75.6 11.2 86.8 4.4 4.4 2.2 2.2
M 20.0 68.9 88.9 11.1 100 0 0 0 0

 

 

 

 

70

household core of a married couple (with or without children). Tables 3.3 and 3.4 show
that there was overall similarity in household type between the black and white samples, in
1880 and 1910. In both time periods, white owners had the lowest incidence of nuclear
households, and were especially numerous in the categories of adult siblings and other kin.
This pattern of adult siblings and other kin forming the core of white households may be
related to the management and transmission of property for the white sample.

Although the same pattern is present for the black sample in 1910, the difference
between black renters and black owners is not as large as the difference between white
renters and white owners. In contrast with white owners, black owners in the 1880 data
set have the highest incidence of nuclear family core structure, with no households formed
by adult siblings or other kin. These core types, however, are represented in the black
owning group in 1910, and in larger percentages than among the black renters. The reason
for this change is not obvious, but property ownership in the black sample, which was
increasing over time although it remained much lower than the white sample, may have
affected household structure. Household type 3, single parent with children, shows little
consistency between the subsamples. In 1880, type 3 households were more common
among whites than blacks for owners and renters, but not for sharecroppers. In 1910 type
3 households were of needy «pal incidence for all groups, except for black owners.

Most households were headed by a man. If an adult man and woman were both in
the household, usually as a married couple, the man was almost always designated the head
of the household by the census taker. The designation of the man as head and the woman
as spouse when both were present is taken as given by the census taker and is here
intended to convey information about joint residence rather than relations of power within
the household. Females did head households in which there were adult male boarders,

and in a few instances were designated the head with the man as spouse. The reason for

71

this designation is not known, but possibilities include the women owning the house or
farm separately. The incidence of households headed by women is presented in Table 3.5.
The most obvious trend is the increase in this variable from 1880 to 1910, for all
subsamples and especially for black renters, for whom it doubled. White owners showed
the highest incidence of female-headed households, both in 1880 and 1910. Given the
prominence of the black female-headed household in sociological analysis, this variable is
explored further below, from the perspective of women of various ages within households.

Table 3.6. presents statistics on numbers of persons and the dependency ratio in
1880, by household types. For the most part, tenure or racial variables are less important
than household type in the variation of the household demographic measures. The
dependency ratio (comers/Workers) is presented in the last column (see Appendix for

age breakdowns), and shows much greater variation within household types.

As mentioned above, additions could be made to households, independent of the
core household type. Tables 3.7-3. 12 present this perspective on household composition,
with only the more common categories shown in the tables. The first type of addition
considered is household extension, or the inclusion of lineal relatives (Tables 3.7 and 3.8).
Most households were not extended, and there is an overall trend, even if weak, toward
decreased extension from 1880 to 1910. The highest incidence of extension in 1880 was
among black renters and in 1910 among black owners. More black households were
extended, although white sharecroppers (together with white renters in 1910) also had a
large incidence of extension, predominantly by the addition of parents of the head or
spouse. Also common, especially for black households, was the inclusion of a daughter
and her children. The ”stem” household, in which a married child resides with the parents,

was very rare in 1880 and more common in 1910, especially for the white sample.

72

Table 3.5. Female-headed households, Vinton, 1880 and 1910 compared.

 

 

 

 

W W
1889 12.19
Black 06 14
rent {11:339-331.}:11219-4321 06 13
W 04 07
Muir: 08 11
rent (H]§§Q-3§,H12]Q=45) 05 07
8 - 910-45 09 15

 

Children living with grandparents, without their parents, was also an important type
of household extension, especially for the black sample. Raper’s study of Green and
Macon Counties, Georgia in the late 1920s and early 1930s revealed a similar pattern; 2.5
percent of white children versus 8.7 percent of black children lived with grandparents
(Raper 1968:71). In 1910 Vinton, grandchildren were most common in female-headed
black households, especially female-headed households that owned their farms. These

households probably had more resources to care for grandchildren.

The category of blended or complex households was tabulated separately in 1910.
This tabulation was not possible for the 1880 data, because of the less detailed information
on marital status. In 1910, complex households, in which at least one partner had been
previously married, had the highest incidence of extension. These households would have
had a larger pool of potential ancestors and descendents (including those from past

marriages) to draw upon.

73

Table 3.6. Demographic measures by household types and race/tenure groups, 1880,
means.

 

 

 

 

 

 

 

Type GroupN total total nuclear consumers workers dratio
persons family family

1-2 blow 19 5.8 4.9 3.2 3.2 3.8 1.25
1-2 blre 161 5.9 5.06 3.2 3.7 3.0 1.26
1-2 blsh 160 5.5 4.9 2.8 3.5 2.8 1.25
1-2 whow 46 6.3 5.2 3.8 4.2 3.6 1.18
1-2 whre 19 5.6 4.9 3.1 3.6 3.1 1.18
1-2 whsh 10 5.3 4.6 3.2 3.6 3.0 1.2
3 blow 2 3.5 2.0 2.5 2.5 2.05 1.25
3 blre 17 5.4 4.2 2.3 3.0 2.06 1.09
3 blsh 22 5.6 4.6 2.5 3.6 2.5 1.44
3 whow 9 4.8 4.1 3.7 3.8 3.2 1.15
3 whre 3 5 4.7 4.0 3.5 3.4 1.03
3 whsh 0 at

4-5 blow 0 at

4-5 blre 2 9 3 6 5.0 4.5 1.09
4-5 blsh 6 5.0 5.0 3.0 4.0 2.7 1.48
4-5 whow 1 4.2 2.5 3.8 3.4 3.2 1.06
4-5 whre 0 2.0 2.0 2.0 1.8 2.0 .9
4-5 whsh 0 m

6 blow 1 4.0 1.0 2.0 2.7 2.3 1.17
6 blre 3 4.3 1.0 2.67 2.3 2.2 1.04
6 blsh 2 2.75 1.0 2.25 2.32 2.2 1.06
6 whow l 5.0 1.0 2.0 2.7 2.0 1.35
6 whre 1 4.0 1.0 3.0 2.7 3.0 .9
6 whsh 2 1.5 1.0 15 1.4 15 .95
7 blow 8

7 We 6 2.3 1.0 1.7 1.8 1.5 1.15
7 blsh 4 1.0 1.0 1.0 1.0 1.0 1.0
7 whow 3 1.5 1.0 1.5 1.5 1.5 1.0
7 whre 2 m

7 whsh 0 m

 

 

74

Table 3.7. Household extension, by race and tenure groups, 1880.

 

 

 

(table giyes pemmLage of households from g_ro_up having specified extension)
none mother parents da., da,husb, grandkids combi- gand-
+ kids + kids alone nations mother
Group N
blow 22 89.2 4.5 13.16
blre 189 79.3 5.3 6.4 1.1 5.9 1.6 .5
blsh 192 82.3 3.8 5.7 2.1 1.0
whow 64 90.6 9.4
whre 26 88.5 3.8 3.8 3.8
whsh 12 83.3 16.7

 

 

Table 3.8. Household extension, by race and tenure groups, 1910.

(table gives 96 of households from each raceltenure group having specified extension)

 

 

 

none parents da + kids child, kid, grandkids grparents
+ spouse alone

GroupN (STEM)
blow 67 82.1 3.0 4.5 1.5 7.5
blre/sh 482 88.2 5.0 1.4 .2 5.0 .2
whow 45 86.7 6.7 2.2 2.2
whre/sh 45 84.4 13.3 2.2
BleuIedHI-ls

blow 26 74.0 3.9 13.1 9.1

blrc/sh 145 82.2 9.3 .8 7.6

 

 

 

75

 

m.w bd— hd— 92. 2 :33

9.: m.m w.m QM 03 on 0.23

 

 

3 S. 3: 3 S. 3 3m $1323
w. m. 3 n. 3 S. 3 3 3 3m «ladders.
m. n. m. 2 m. 3 Z 2 3w olwl_..e._.n.
3. 3. 3. 23 MN. 32..
9.80
32.9: as: 8:
.822. .5 :93: .85 «Ex new 37.
N. -568 0.25 E3 2:38 83: enemas. .8595 .837. .1 :2: 8.3.9. .8595 0:0:

 

€28.35 358% 9.32 9.2m 8:538.— zoee Go 33:38: ..o ems—520; 32» 035V
.3“: .385 38 can 83:8 .3 62385 20:38: .a.m 033.

 

 

76

Table 3.10. Household expansion, by tenure and race groups, 1910.

(table fives 96 of households of each race/tenure m having specified expansion)
none siblings sis+ bro, sis, niece cousins combination
kids wife, husb., or
kids kids neph

 

913312.131

1219.15.62 92.5 3.0 4.5

W 91.9 2.3 1.5 3.5 .2 .4
m 91 1 6 7 2 2

 

M 95.6 4.4

Table 3.11. Household augmentation, by race and tenure groups, 1880.
7 (table gives percentage of households of each gloup having specified augm_entation)

 

 

 

 

.8. m mans: m mm
mm manta
m 77.3 4.5 9.1 9.1
1213.182 75.0 3.6 19.1 2.1
mg: 87.0 6.7 5.7 .5
M 60.9 6.3 29.8 3.1
m 76.9 0.0 11.5 7.7 3.3
mm 83.3 0.0 16.7

 

 

 

77

Table 3.12. Household augmentation, by race and tenure groups, 1910.

(table gives 96 of households in each Mmqum)
nan:

 

 

 

91911101

1219151 85.1 15.0
121M 95 4.9
W 89.0 10.095
W 91.1 8.9
aha-£86.35 80.0 19.9
W 70.0 30.0

 

Extension covers lineal descendents and ancestors. Lateral relatives of the
household core are classified under the expansion category. Generally speaking,
household expansion was even less common than household extension, except for white
owners and white sharecroppers in 1880 (Tables 3.9 and 3.10). The white sample clearly
showed more household expansion in 1880, especially by the inclusion of brothers and
sisters, compared to the black sample or the white sample of 1910. The presence of nieces
and nephews in households, without their parents, was more common for the black

samples, in both 1880 and 1910.

The final addition to households is non-relatives, or augmentation (Tables 3.11 and
3.12). With the exception in 1880 of one persons noted as a ”friend,” perhaps to indicate
a temporary visitor, non-relatives are described as boarders or servants. The determination
of someone as a boarder or servant is accepted directly as given by the census enumerator,
presumably from information provided by household members. It is possible that some
boarders and servants were related to the head of the household and his/her family, but not
closely enough for the residents or census enumerators to describe the relationship. The

overall trend from 1880 to 1910 is for less augmentation over time (see also Figures 3.5

78

and 3.6), but with different trends for boarders versus servants. The incidence of boarders
in households increased from 1880 to 1910, while the incidence of servants in households
decreased to zero. The complete absence of servants in 1910 is somewhat arrious. Some
of those persons listed as boarders in the 1910 census of Vinton were listed with the
occupation of ”laborer.” Some boarders may have worked for the household in which they
resided, and perhaps would have been more accurately described as servants. However,
the designations given by the census have been followed. White households had more
servants in 1880 than did black households, but not by much. Black households had more
boarders in 1880 than did white households, while in 1910 white households had more
boarders than did black households. Female heads were tabulated separately in 1910, and
showed a high incidence of boarding, especially for the white sample.

W:

The above discussion looks at all households at one time, regardless of the age of
the head or the ages of children, or other factors related to the life cycle of the household.
However, the demographic circumstances of households change as they mature, and these
changes could have important effects on household composition and the economic welfare
of the household. Table 3.13 shows basic household composition for 1880 Vinton
households, subdivided by household stage. As one would expect, the number of nuclear
family members in the household changes dramatically according to household stage. The
size of households peaks in stage 3, which is a middle stage when some (but not all)
children are over fourteen years of age. However, the number of consumers and workers
does not peak until stage 4, which is a slightly more mature stage. These differences are
illustrated in Figures 3.7, 3.8, 3.9, and 3.10. Each clustering of vertical bars in a figure is
one separate tenure and racial grouping. The series of bars within the cluster represent that
tenure and racial grouping’s changes over household time. That is, the clusters capture

79

changes over the household life cycle. The dependency ratio, or the consumer/worker
ratio, which is more ”favorable” the lower the value, is highest in stages 2 and 3. This is
when there are the most children in the household. The dependency ratio begins to decline
in stage 4 and declines further in stage 5, the most mature stage (Figure 3.11). Within the
stages, the consumer/worker ratio and other variables are fairly consistent across tenure and
racial groups, although there are minor variations according to each grouping (see Figures
3.7-3. 1 1).

Not only do overall household numbers vary by household stage, but the types of
additions to households, extension, expansion, or augmentation, vary by household life
cycle stage as well. The category of household extension is not independent of household
stage. This is because lineal relatives are linked in time with the core household persons.
For example, grown children are not available as adult additions to their parent’s household
until the original household has reached a sufficient stage for the children to be adults. Or,
parents are generally not available to live with adult children until all of their own children,
or the children in question’s siblings, haVe left the original household.

However, other categories of household addition, like expansion with lateral
relatives, and especially augmentation with non-relatives, are less intimately connected with
the household life stage. This is not to say that household stage does not affect these
aspects of household composition, only that the effect is less closely tied to the household’s
life cycle stage. Table 3.14 presents summary measures for household type,

80

Table 3.13. Demographic measures by stages and race/tenure groups, 1880, expressed as
means.

 

 

 

 

 

 

 

m Group N total total nuclear consumes workers dratio
Stage persons family family consumers/workers
I blow 0 m

1 blre 10 2.7 1.9 2.7 2 3 2 3 99
l blsh 11 2.4 2.0 2.2 2 0 2 0 99
1 whow 2 4.0 3.0 4.0 3 6 3 5 1 02
1 whre 1 1.0 1.0 1.0 1 0 l 0 l 0
1 whsh 2 3.5 2 0 2.0 2 5 l 8 l 35
2 blow 8 5.4 4.9 2.25 3 2 2 5 1.3
2 blre 70 5.9 5.3 2.5 3 3 2 5 1.3
2 blsh 89 5.4 4.9 2.3 3 l 2 4 1.3
2 whow 20 6.2 4 9 3.0 3 6 3 0 1.2
2 whre 11 6.0 51 2.8 3 5 2 8 1.2
2 whsh 4 6.7 5 7 1.1 32 84 .29
3 blow 4 7.75 4.75 4.25 4.8 3.6 1.3
3 blre 38 7.6 6.8 3.6 4.4 3.2 1.4
3 blsh 35 7.1 6.2 3.3 4.4 3.3 1.3
3 whow 12 7.3 6.5 4.4 4.9 3.9 1.27
3 whre 1 6.0 6.0 3.0 3.8 3.0 1.26
3 whsh 0 m

4 blow 3 7 3 7.0 4.7 5 3 3 8 1.4
4 blre 21 7 l 5.7 4.0 4 7 3 6 1.3
4 blsh 22 6 5 5.9 3.8 4 5 3 5 1.3
4 whow 8 6 4 5.5 4.9 5 1 4 7 1.13
4 whre 4 6 25 6.25 3.5 4 12 3 45 1.2
4 whsh 2 6 5 6.5 5.0 5 4 4 65 1.17
5 blow 6 37 3.0 28 29 26 1.1
5 blre 39 43 2.8 34 32 28 1.1
5 blsh 23 3 7 2.8 31 31 2 9 1.0
5 whow 11 42 3.8 3.9 37 34 1.1
5 whre 5 4 0 3.2 3.8 3 4 3 4 1.0
5 whsh 2 3 0 3.0 3.0 2 8 2 7 1.03

 

 

81

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o o oo— o oo_ o o oo_ o o o o oo_ 5...?
on o on o oo_ 3 o 2. o o o mm 2. 9:?
o. o ow o_ ow o o o2 o o o on on 32.3
vo o 3 o on 2 3 mm o o o o oo_ 5....
om no 2. No 5 mm No an no no o S a» £3
2 o 3 o oo— mm 2 on o o o mm 8 >63 3
o o oo_ o o3 o o oo_ o o o o oo_ 5:3
o o oo_ o oo~ o o oo— o o o on ow 9&3
em 3 en 3 ma o vo 3 o o o 3 mm 32.3
vo no mm .o no 2 no 2. o o o en 2. £5
2 3 2. o no 2 no on o o _o E v» 23
o 2 mm mm R 2 o mm o o o o oo— >55 fl
om o ow ov ov o ov oo o o o o oo_ 6:3
o o 3 mm 2. o mo 3 mo o o o 3 on?
ma oo an 3 w o 2 om o o no o no 333
no we mm oo 3 o oo oo o o o no vo an...
3 no me S on o : 5 8 o o no no 25
2 o 8 o 2: o o 8. o c o c 82 .63 «a
_ g
11mm? mm.“ .stm & fld g
duo..." .83 28: madam use: 3 malea mg a a mama diam madman a
cows—8 =< coma—adam— eomncofim Boson—o: .

Acme...“ 20:33: .3 .25 Eon—03°: @056on 5 8a .2. 0&8... 55% E 8.23:0: he “58qu 82m 335

.oww- .eoEmooEoo 20:83: .35 oSaH

 

 

I Stage 1

I Stage 2
5 Stage 3

W Stage 4

 

 

 

 

 

 

 

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Total persons, by life stage.

Figure 3.7.

I Stage I

I Stage 2

5 Stage 3

83

1 Stage 4

El Stage 5

 

 

-«-<.

TEE;

 

 

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Figure 3.8. Total family, by life stage.

I Stage 1

I Stage 2

E Stage 3

a Stage 4

84

El Stage 5

 

 

 

 

 

 

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Figure 3.9. Consumers by hh life stage.

I Stage 1

I Stage 2

E Stage 3

M Stage 4

Cl Stage 5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Figure 3.10. Workers by hh life stage.

I Stage 1

E Stage 4

86

1:] Stage '5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Dependency ratio, by hh life stage.

Figure 3.11.

87

extension, expansion, and augmentation by three household life cycle stages, young,
middle, and mature. Emansion is generally slightly higher in the middle stage, although
the presence of siblings is most common in the early stage, perhaps before the siblings go
into households of their own. The incidence of adult siblings of the household head or
spouse in the household continues to be high in the middle stage of the household life cycle
for black owners. The relationship of augmentation to household life cycle is not clear.
Servants are more common than boarders at every stage. Boarders are especially rare in the

mature stage.

2 'I' 'Il' II 11!-

In contrast to the data presented above, which is from the perspective of the
household itself, Vinton household composition can be viewed from another perspective,
that of individuals within households. It is possible to look at the numbers of individuals
within different types of households. This produces similar results as the household
classification discussed above, except when certain types of households are smaller or
larger than average. For example, solitary household make up 10.4 percent of black
owning households in 1910. But since these households have the smallest number of
persons (by definition one person), less than 10 percent of persons in black owning
households lived in solitary households.

More useful is the relative importance of various positions within households.
Position within households is a standard data category provided by the census taker,
defined by the relation of each member to the head of household. The possible categories
are head of household, spouse, child of head, parent of head, niece or nephew, cousin,
sibling, aunt or uncle, grandparent, grandchild, boarder, and servant. This perspective
teases out the individual experience and, like the household life cycle analysis presented

88

above, helps overcome the static nature of household type classifications. The designation

of the man as head and the woman as spouse was provided by the census enumerator.

The position of an individual within a household changes dramatically as people
age, beginning life and growing up as children (or boarders) to eventually head their own
households or reside in households with others as spouses, relatives, boarders, or
servants. Because of this process, information on household is presented by various age
categories. Tables 3.15-3.17 present this perspective, with data given separately for men
and women and by various racial and tenure subsamples. For the most part, the black and
white samples are similar, as was noted in the discussion of household structure.
Important differences exist, however, and will be discussed at this point. Looking first at
1880 (Table 3.15), the most striking difference is the frequency of the relationship ”child to
head”. For both men and women, nearly double the percentage of whites compared to
blacks reside in the household of their parent(s). This difference is quite dramatic in the 17
to 24 age group, but persists through all ages. Except in the 17 to 24 age group, this

position is more common for women, both black and white, than men.

The frequency of individuals residing in households in which a sibling is the head
is more common for white men and women, especially in the young age brackets. The
percentage of persons of all ages who were boarders, servants, or other kin was very
similar for all subsamples. White women were least frequently servants or boarders in
other households, especially at older ages. Although it is not broken out in Table 3.15,
some of those persons related to the head as ”other kin” for the black sample were
grandchildren or grandparents of the household head, which suggests the importance of the
alternate generation bond in the black population. Grandparents and grandchildren of the
household head were less common household members among the white sample, where

the ”other kin” category was more heavily composed of cousins, aunts, and uncles.

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 o c o : 8 8 o o o o 8 11.5.58 .4
8 8 8 .o 8 8 8 8 8 8 8 8 5:8 a
8 o 8 8 8 o 8 o o o o 8 8.56m .3.
8 8 8 8 8 8 8 8 .o 8 8 8 88ml! 3.
.o 8 o o o 8 8 e. o 8 o .o .5. :50 Hal
8 o o no 8 8 - .o 8 no 8 o 8 =2.2.6 q
8 8 o .. 8 a. 8 o 8 8 8 a. 8.3% fit
8 8 o 8 Z. 8 8 o o 8 8 8 8.3%. H:
o o o o o o 8 8 8 o o o .Icaaaui fl
8... o .8 8 o o 8 8 m. 8 c o .15.... .I..I
8 o 8 : 8 o... 8 o t 8 : I 8 2. as
a. o .8 .o 8 8 m. o 8 8 8 8 .2. 3|
8 o o o o 8 cm 8 a 8 8 8 088m .131
.8 o o 8 8 no 3 8 8 8 E 5. 38mm 31
8 8.. 8 E 8 : 8 m : t 8 8 Pas-:- fl.
8 8 8 .o 8 8 n. 8 8 a. : 8 88:1- .lal
88. a. 8 8. o: 8.. 82 am 28 88 8 new 3 z
8. n 8 on 8 8 o: e S 8 8 a. e... .z
.|.| I I I l l I 1.1 Ill 1.1 I ll
.7. +8 8.9. 8.8 8-8 8-: .3 .8 8-8 3:8 8-8 8-: 5:86.. mg
0 Na .52 mmsmfl em up €2.83

Ame—9:8 8...? e5. :83 8.. £658.. 223.6: 058% E 96..» one 58 no ems—820.— 8>mm 038V

.8... .823 8.2.832... .82. s 8.823.526 n.._.a_e......3. .m. .m 6.3

 

90

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o o o o o o o o o o o o |.§Em ..3
o o o o o o o o o o o a .IIEEm E:
o. 8 8 8 E o. 8 o o 8 o 8 38m. :3
8 8 8 8 8 8 8 N. 8 8 8 8 3.8m H:
o o o o o o m... o 8 o o 8 5.22.6 .5:
8 8 o 8.8 8.8 8 8 8 .o 8 88 8 .5. .950 H.
8 o 8 8 o 8 8 8 8 8 8 8 8.35 ..3
8.8 o o o 8.8 8 N... 8 8 o 8 8 8.3% fl.
8 8 o o o o 8 8 8 o o o .ISE. .5
8.8 8 2. o o o 8 8 8 .o 88 o .53.. a
8 o o. o z 8 : o o 8 8 8 Eu .5
m. 8 he 8 o. 8 : o .o 8 8 8 2.5. H.
o o o o o o 8 o 8 8 8 8 3.54.8 .5: I
8.8 8 ho o o o 8 8 .8 8 8 8 8818 El
8 8 8 8 8 m. 8 8 8 8 o o flaw: .3
8 8 8 8 8 8 8. 8. 8 : .. 8 as: a

8 8. 8. 8: 88 8 8. 8. 8. 88 3.2

o 8 8 .8 8 I m 8 .8 8 l8 €22
.la +|8 8-8 388 8.8 8.: fl +|8 8-8 3:8 88 8-: -...o-.....8... 820

 

=0.— oa £52

 

=

 

o a a 5:53

Ravi—Ea 823 v5 “—83 8.. :83qu 22.8.5: 058% 5 95..» can :98 he $859.2. 83» 03.5

.2 a. ...o....> 6.9.8888 88.. 2 8.322858 8.8.5.328285. .8. .8 938

 

91

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

n. o N. S 8. I 8 o o c o a mafia a.
.8 3 3 o .8 .8 8 o o 8 c o g a
3 8 no 8 .8 no 3 N. a... no 8 no god 8.4.
S .8 .8 o n. 8. 2 N. .. o N. o 8.458.. Jamal

o c o o o o c o o o o o 8.42.40 mam.-

.. o a o o o .o o 2 o o o 3 final
8... 8 o a... :o .o 8 8 8 .o o .8 slag H.8-
8 o o o o o 8 N. o o .8 8 add 83..
n o o o .8 o I 8 o o .8 o S 8...... .8. «88.
8 o .8 o o o z. o. 3 o S o a EN.
8... o o o :o no 8 8 .8 o 8 .o 33% m8...

8 o o o o o .8 o o o 8. o 3.3.8 Howl
z. 2. c o c o 2 2 .. o o c .5... mum.
3 3. c c o o. c o o c o c g a
e... 8 a... o c o .8 on c. 8 8... o a...“ ma...

8 c o o o o 8 c .8 o o o g .33..
N. c o o 8. .8 8. o c c S 8. a m...”
.8 c o 2 .8 8 .8 o c 3 8 8 3.4.... an
N. o a... 8 .8 .8 .. o o 8 8 on «a Hal
.8 c o .8 .8 .8 2 o .8 N. .8 8 .... final

8 o o o o o E o 2- 8 2. .8 83w .42...”

o o o o o c .8 o 8 z. .8 8 an a..."
«.o c o a... o o 8 8 8 .2- 2. 8 mafia an:
.o o .8 o o o .8 8 8 .8 2. 2 .oflHflLm final
8 3 8 2 .8 N. 8 8 .. .8 o a man EN.
8 .8 S 8 o. N. 8. o 8 3 c c an ”Hal
8 2 2. 2 2 .n 8. 8 8 N. I no a...- man
.8 8. 2. .8 .8 - 3 an o N. .8 8. o 8mm Hal
«.8 8... [8-8 j... 8-8 I8... I.... file 8.8 3.8 8-3 INI: é flaw“.- 8.8..

2% E03

A 3.955 05.5. :98 3.. :82. o. coca—2. .3582. 22.8.5: Sets, a. 95.5 on“ :08 ..0 098.880.. 82m 038.

.22.... 3 .2... ..§...> .858. 22.3.5: :8 2......-

 

92

Women, both black and white, were much more likely to live in the householch of their
children. Although these patterns are may be influenced by longer life expectancy for
women, the difference between men and women also exists at young ages. There were in

fact no white men in this position in 1880.

The most obvious difference between men and women’s position within
households was the incidence of the headship and spouse position. Thus the most
common household position for all women was spouse, and the most common position for
all men was head of household. Black men were slightly more likely than white men to be
household heads, and black women were slightly more likely to be spouses. This
difference was especially strong in the 17 to 24 age group, which suggests an earlier age at
marriage for blacks (discussed below). Female headship was rare compared to male
headship, but was the third most common position for women, after spouse and parent in
child’s household. Overall, a higher frequency of female headship was indicated for black
women. For both groups, female headship was lowest at the youngest ages, and increased
with age. As mentioned before, the few cases of black men as spouses and not heads of
house is intriguing. Note that this status occurs only in the three earliest age groups in
l 880.

Table 3.16 presents the same information as Table 3.15, only for 1910 instead of
1880. Many of the patterns described for 1880 are still present. These include the higher
incidence of women than men in the household of their children for the black sample, the
higher incidence of white individuals residing in a sibling’s household, and the low
occurrence of white women as boarders. One major difference between 1880 and 1910
was the loss of the servant category, which, as suggested above, maybe related to
differences in the census taker's views or classifications. Correspondingly, boarders were
more prevalent in 1910 than they were in 1880.

93

The most dramatic difference between blacks and whites, the higher frequency of
white adults residing with parents, continued, although the degree of divergence by race
was stronger for men than for women. As in 1880, this difference in 1910 was strongest
at the earlier ages, suggesting blacks left the parental household at an earlier age. A very
high percentage of black men age 17-24 were heads of their own households. The pattern
of men as heads of households and women as spouses continued as well. Again black
men were more likely to be household heads than white men. However the pattern
observed in 1880, in which black women were more likely than white women to be
spouses, was not observed in 1910. In fact, for the 25-34, 35-44, and 45-64 age groups
in 1910, white women were more commonly spouses than black women, with black
women more commonly spouses at the 17-24 and 65+ groups. The frequency with which
black women and white women were spouses increased dramatically from 1880 to 1910,
from 50 percent (white) and 57 percent (black) in 1880 to 65 percent (for each group) in
1910. This increase is present at all age groups but is especially high for white women in
the 17-24 group, where it represents a jump of 120 percent, suggesting a decline in age at
marriage for white women. Other trends from 1880 to 1910 include a slight decline in the
incidence of living with siblings. This could easily be caused by a slight decrease in age at
marriage, assuming marriage meant the start of a new household, which it generally has in
the West (Hajnal 1965,1982). The percentages of persons living with other kin also
declined over time, except for black males.

Since the percent of women who were spouses declined dramatically from the 45-
64 to 65+ age groups, for both 1880 and 1910, a closer look was taken at 65+ women and
their positions in households. In fact, for both 1880 and 1910, no white women 65+ were
living with a spouse, compared to 29 percent of black women in 1880 and 41 percent in
1910. This pattern suggests longer life expectancy for women than men. The model life

94

tables used below in the fertility analysis suggest that the life emectancy of women age 45
was two years longer than that of men age 45, for blacks and whites. All of these women
list their marital status as widowed. Some of these women who were 65+ and were not
spouses headed their own households. This situation was more common in the white
sample, for both 1880 and 1910. However, anywhere from 30 percent to 60 percent of
women who were not spouses resided in their children’s households, depending on race
and time period. Seventeen percent of white women 65+ lived with other kin in 1880, but
none did in 1910. About five percent of black women 65+ lived with other kin, in 1880
and 1910.

The number of children who were surviving when a women reached age 65 seems
to have played an important role in determining her residence. The household positions of
women 65+ in 1910 were recalculated according to the number of children the woman had
surviving (these data were not available for 1880). The incidence of widowed women
residing with a child went from 0 percent if one child was surviving to 100 percent if three
children were surviving. The only widowed women living as boarder for whom

information on number of children surviving was available had one child surviving.

Wm

Differences in household composition, processes of household formation, and
fertility by tenure are important aspects of this research. The theoretical perspectives
outlined in Chapters 1 and 2 suggest that there will be differences between owners, renters,
and sharecroppers, stemming from variable constraints on household formation linked to
access to productive resources. To test for these differences, positions in households were

broken down according to tenure. This data is presented in Table 3.17.

9 5

As with the analysis of household composition and type presented above, the
similarities between renters/sharecroppers and owners are sometimes stronger than
similarities between blacks and whites. This is the case for the categories of siblings,
other kin, and boarders. This suggests that different cultural heritages may affect
household composition more than tenure. However, differences by tenure are important
for the categories of child, spouse, and head. For example, owning households, both
black and white, were much more likely to have children age 17 to 24 living with their
parents. While 30 percent of 17 to 24 year old women were living in their parents’ house
in the subsample of black renting households, 83 percent of such women in the black
owning subsample were living with their parents. Similarly, while 43 percent of white 17
to 24 year old women were living with parents in the submmple of white renting
households, 67 percent of such women were living with parents in the white owning
households (Table 3.17). This difference between renters and owners continues through
all age groups, although of course the incidence of adult children living at home dropped
dramatically as the children aged. It is an important feature of the tenancy stratification that
owners of farms usually cultivated more acreage and would have been able to make better
use of and provide better support for the labors of their children than would renters. They
also would have had more material wealth to pass on to children, which may have changed
the way marriage was viewed.

The same pattern of differential residence in parents’ household by tenure exists for
men. Here the black and the white samples are very similar, with about 30 percent more
17 to 24 year old men still residing with parents among owning households compared to
renting households (Table 3.17). A small portion of sons and daughters living at home
were already married and, with their spouse, perhaps in line to take over the family farm.
This can be seen most easily in Table 3.8, in the higher incidence of married children living
with parents (stem family form) for households which owned their farms. This pattern

96

holds by both tenure and ethnic breakdowns. However, this household form is relatively

rare, even for white owners for whom it was more common in 1910.

The suggestion that these differences in the incidence of persons living with parents
is related to differences in age at marriage and the rate of new household formation is
supported by differences in the positions of head and spouse, and not, for example, in
categories such as other kin or boarders. For the younger age groups l7-24 and 25-34, a
much higher percentage of black women were spouses in renting households than in
owning households (57 percent compared to 17 percent for the 17-24 age group, and 80
I percent compared to 62 percent for the 25-34 age group). The same pattern is also present
for white women. Similarly, a much higher percentage of black men in the youngest age
groups were heads of households in the renting subsample than in the owning subsample.
This difference by tenure was not supported in the white male sample, although it was
supported in the white female subsample. Young adult white men were less frequently
residing with parents in renting households. However, instead of being heads of their own
households, as was the pattern for young adult black males, young white males, especially
in renting households, were more commonly boarders or living in a sibling’s household.
The pattern of higher incidence of headship for renters does hold for white males of the
next age group, 25-34, so that the general pattern may hold for white men but just be
delayed by an average later age at marriage and a tendency to live with (presumably older)

siblings before their own marriage.

Other differences by tenure include a higher incidence of persons as boarders for
the black owning subsample. This is in contrast to an opposite pattern for white Vinton
residents. The difference may reflect a lessened ability of the black renting/sharecropping
subsample to take in boarders, given that their houses were often very small and owned by
a landlord. Black owners likely had more room, greater residential stability, and greater

97

control over household composition. Conversely, it is possible that all white households
had the space to take in boarders or extra family members, especially compared to black
renters/sharecroppers, but that it was white renters who felt a stronger need for the extra
household income’labor that additions to the household fumished. Another difference
between renters and owners is the higher frequency of female-headed households for black
renters. The same is true for the white sample except in the 45-64 age group, which exerts

a strong effect on the overall distribution.

CHAPTER 4
MARRIAGE AND FERTILITY

Matias

Chapter 3 showed that most Vinton households in 1880 or 1910 were composed of
a nuclear family. This chapter presents an analysis of marriage and fertility, the two most
important processes through which nuclear family households are formed and develop over
time, for the Vinton population. Because of the importance of rural and urban differentials
in theories about fertility, a sample from the nearby county seat, West Point, Mississippi, is
added for the fertility analysis. The fertility analysis, which utilizes a variety of measures
of the level and nature of fertility, is followed by an indirect estimation of illegitirnacy for
the Vinton and the West Point samples The broad reach of this entire study, encompassing
not only fertility but marriage patterns and household structure, makes it possible to give a

broader contextualization to illegitimate births.
Mgrigl Sms in yigtgg

As suggested by the high frequency of nuclear family households and the spouse
position within households in Chapter 3, one of the most universal experiences for turn of
the century women was marriage. Table 4.1 presents data on marital status by race in 1880
and by race and tenure in 1910. Detailed analysis of marriage was reserved for the 1910

sample, since the 1910 census gave better information on marital status.

98

99

 

 

 

 

 

 

 

 

88 88 88 88 88 88 88 88 finale
8. 8. 8 88 8 .8 8. .8 seal
8. 8. .. 88 88 88 88 88 s. 8 8
8. 8. 8. 88 88 .8 88 88 wall
88 .. 88 88 88 8. 88 88 I.=.2||les
8. .. 8. 88 88 8. 8. 88 g
8. 88 8. 88 88 88 .8 .8 .. 888
8. 8. 8. 88 88 88 8. 88 833.8888
888 .888. 888 .888. 88. .888 888.888 888 .888. 888 .888. 88. .8.8 888 .888 .8 .8.8. .888. .z
.8. .88. .8. .88. 88 .88 88.88 88. .88. 88. .88. .8 88 88 .88 .5 .8.8. .888. .z
888 =8 888 =8 88-88 88-8. 888 =8 888 =8 88-88 88-8. malpfi
3303330988— 0—qcma «coohom 330232—0988— U—M-umu ~c00uoh

ZmEO.‘

ZN:

803023 .0 29:8 8. .85 use.» :08 no 03.8808. 32m 058.

888858 8. 8. 888 888. .8358 3.88:. ...8 8.888

 

100

Table 4.1 shows the percentage of individuals who were either single or widowed.
Since divorce was very rare in the sample population, for most purposes it can be assumed
that persons not widowed or single were married. The percentage of black men who were
single decremed slightly from 1880 to 1910, while the percent that were widowed
increased by nearly the same amount. The percentage of white men who were single also
decreased over the 30 year period. In both time periods, a higher percentage of white men
than black men were single and a higher percentage of white men were widowed. Owners
showed a higher proportion of widows than did renters, for both blacks and whites.
Owners showed a higher proportion of single men for the white sample. Many of these
e-e tendencies were repeated for Vinton women. The percentage that was single
decreased over age, while the percentage widowed increased. Owners showed a higher
percentage of women single, and a lower percentage of women who were widowed. Thus
it appears that while the state of marriage became slightly more common from 1880 to
1910, marital disruption may also have increased.

An analysis of differences in marital status between the various age groups, by
tenure, supports the earlier findings regarding position in household. That is, for the
younger age groups 17-24 or 25-29, higher proportions of owners than
renters/sharecroppers were single. The differences are most dramatic for the black sample,
again suggesting earlier age at marriage for black renters/sharecroppers. Higher
proportions of women were widowed and lower proportions of women were single, for all
age groups. Widowhood was especially prevalent among black women in 1910; almost

one quarter were widowed.

101

eho

The 1910 census included detailed information on marital status, and the analysis of
marital status was expanded accordingly. Serial marriages were tabulated by their number,
and all marriage statuses, not just the percent single as in Table 4.1, were analyzed
separately by the age of the person. Some analysis was conducted for men, but more
detailed analysis was conducted for women. This is because of the increased importance of
women’s marital status to a population’s reproduction rate, given the link between a
woman’s marriage and fertility. Black women were also subdivided by the tenure of the
household in which they resided. This subdivision by tenure was not possible for the

white sample due to its smaller size.

Table 4.2 presents the results of the breakdown of marital status by age. Steadily
decreasing percentages of single men and women are seen among older age groups, for all
subsamples. This is especially true for black women, who by age 65 had all been married
at least once. This universality of marriage is less strong for white women, although the
sample of white women is smaller and therefore not as reliable as the larger black sample.
The large percentage (20 percent) of white women 65 years and older who were single is
undoubtedly a sampling bias.

Although it has been shown that the majority of adult women married, and that
most households had at their core a married couple, marital statuses varied significantly
according to racial and tenure groupings. Higher percentages of black than white women
were serially married, i.e. in second, third, or fourth marriages, for all age groups. More
than half of black women age 65 or over who were married were in a second or later
marriage. Although there were in the sample no married white women age 65 or over to
compare with this black sample, comparison is possible for the next younger age group,

102

Table 4.2. Percentage of age and tenure groups in various marital statuses, Vinton, 1910.

 

 

 

Group N 1st 2rd 3rd 4th single divorced widowed
marriage man' man nm
Bl women, 17-24, all 188 55.9 2.1 .5 0 32.5 0 9
renting hh 58.1 2.4 .6 0 30.5 0 8.4
owning h 42.1 0 0 0 57.9 0 0
Bl women, 25,-34, all 186 66.7 10.2 1.1 0 6.5 0 15.6
renting h 68.8 11.3 .6 0 5.6 0 13.8
owning hh 52.6 5.3 0 0 15.8 0 26.3
B1 women, 35-44, all 143 58.7 18.9 2.8 0 2 1 7 16.8
renting hh 58.0 20.5 2.7 0 2 7 9 15.2
owning hh 75.0 14.3 3.6 0 0 0 7.1
Bl women, 45-64, all 137 39.4 21.2 3.6 0 .7 1.5 33.6
renting hh 39.8 19.4 4.6 0 0 1.9 34.3
owning hh 37.5 37.5 0 0 3.1 0 21.9
B1 women, 65 +, all 42 19 19 2.4 2.4 0 2.4 54.8
renting hh 22.6 12.9 3.2 3.2 0 0 58.1
owning hh 16.7 50 0 0 0 0 33.3
Wh women, 17-24, all 28 53.6 0 0 0 46.4 0 0
Wh women, 25-34, all 32 78.1 9.4 0 0 9.4 0 3.1
Wh women, 35-44, all 32 93.8 0 0 0 3.1 0 3.1
Wh women, 45-64, all 30 60 3.3 0 0 10 0 26.7
Wh women, 65 +, all 5 0 0 0 0 20 0 80
Bl women, all ages, rent 568 55.7 12.1 1.9 .2 10.9 .5 18.7
Bl women, all ages, own 105 50.5 19.0 .9 0 14.4 0 15.2
Wh women, all ages, rent 58 70.7 3.4 0 0 12.1 0 13.8
Wh women, all ages, own 68 63.2 1.5 0 0 25.0 0 10.3
Wh men, all ages 149 57.7 2.0 .7 0 32.9 0 6.7
B1 men, all ages 628 52.6 20.5 3.5 .6 16.1 0 6.7

 

 

103

age 45 to 64. Only 61 percent of black married women age 45 to 64 were in their first

marriage, compared to 95 percent of white married women.

Table 4.2 also provides comparative information on serial marriage for men in
Vinton. Black men were more married than any other group, both in terms of a larger
percent being married in 1910 (76 percent compared to 61 percent ofwhite men or 68
percent of black women), but also in that 25 percent of black men were in a second or later
marriage. Only 3.7 percent of white men were in second or later marriages. The higher
incidence of remarriage for men (see also Farley 1970: 136-137), which resulted in low
widowhood, suggests the importance of women in the rural household, as does the fact
that among single-parent households, 93 percent of black and 95 percent of white
households were female-headed. Men especially did not head households without a

spouse when children were present.

These patterns of marriage and remarriage may be affected by the size of the
population of potential spouses. Vinton sex ratios are 89 (males per 100 females) for the
black adult population in 1880, 90 for the black adult population in 1910, 102 for the white
adult population in 1880 and 1 18 for the white adult population in 1910. The relative
shortage of black men would probably make remarriage more difficult for black women.
The extent to which this low sex ratio is caused by higher mortality among black males,
versus selective outrnigration, is not known, although higher male mortality has been
suggested in other studies (see also Farley 1970; Pleck 1979; Furstenburg et. a1. 1978;
Gutman 1976). The relative shortage of women in the white population may have
contributed to the higher rates of widowhood for white men, compared to black men.

There is no consistent pattern or relationship between tenure and serial marriage.

Higher proportions of black women in the age groups 17-24, 25-34, and 35-44 were

104

serhlly married in renting households. In the older age groups, serial marriage was more
common for black women in owning households. The white sample is too small to
subdivide by age. But comparing the entire sample of white women with the entire sample
of black women, not subdivided by age, shows that serial marriage was more common
among white women in renting households. Conversely, serial marriage was more
common among black women in owning households. It should be pointed out that those
groups which had a relatively low proportion of persons in serial marriages, such as
women age 65+ in black renting households, did not necessarily show increased
proportions of women in first marriages. Instead, they often showed a greater percentage
of persons in the widowed category. It is impossible to determine from the census data

how many times widowed persons had been married.

As a result of serial marriage, about one fourth of all black married households
included at least one spouse who had been previously married. Twenty-three percent of all
persons lived in such households. The multiplicity of sumames of children in many
households suggests that children from first and subsequent marriages often resided
together, with their common mother. Although stepfather relationships were probably
more common than those of stepmothers, the frequency of serial marriage undoubtedly
meant that some women cared for children who were not their natural offspring.
Households containing serially married adults did not contain larger numbers of children
than households of first unions. The mean number of persons under age 17 was two, both
in first marriage households and in those where at least once spouse had been married
before. However, the previous analysis of household composition suggested that complex
households had the highest incidence of non-nuclear relatives. It is unlikely that all
relations of affinity disappeared with the breaking of a marriage. The data set contained
several households in which a daughter-in-law or son-in-law continued to reside with her

or his spouse’s family after the death of the connecting spouse.

105

Comparative historical material on marital status, and especially on serial marriage,
is not common. One investigation providing such information is Johnson’s (1934) study
of 612 black families in Macon County, Georgia, in the early 19303. This work revealed
that 43 percent of marriages involved a spouse who had been previously married,
compared to the present finding of 23 percent of such marriages in Vinton, Mississippi, in
1910. A study conducted in Rowanty township, Dinwiddie County, Virginia, in 1878,
documented that over 50 percent of marriages contained at least one spouse who had been
previously married, and that 32 percent of all households contained persons related through
step relatiom (Manfra and Dyksta 1985). The fact that the Rowanty County study
population was 32 years younger than the present study population from Mississippi may
explain this difference. The Rowanty County sample should have included a much larger
percentage of persons who had experienced marital disruptions because of slavery,

compared to the present sample.

AseatEirstMarriase

Many of the above analyses have touched indirectly on age at marriage, and have
suggested differences in this variable by race, sex, and tenure. Age at marriage is a very
important variable, since it largely determines the rate of new household formation, and has
a large influence on fertility. Because of this importance, direct analysis of age at marriage
was undertaken. This analysis could only be done for the 1910 population because the
census in 1910 provided the number of years married as well as whether this marriage was
a first, second, third, or fourth marriage. The latter piece of information allowed the

analysis to be confrned to those in their first marriage.

106

The first step was to take the number of years married, for those persons in a first
marriage, and subtract it from their age. This gave an age at first marriage for each person.
Although some error is inevitable without exact dates of marriage and of birth, errors of
more tlnn one year are unlikely. The next step in the analysis was to calculate a simple
mean age at first marriage, for sex, race, and tenure subsamples. These results are
presented in Table 4.3. They support the interpretation suggested by the analysis of
household structure above. Except for white women, renters married earlier than owners.

Women married earlier than men. Differences by race are not as consistent.

Small differences in age at marriage are very important, and a mean can be
significantly influenced by unusual cases. Therefore, a more reliable measure, such as the
singulate mean age at marriage, was also calculated. Based upon the proportions single
for any age group and calculated following Hajnal (1953), singulate mean age at marriage
estimates the mean number of years lived by a cohort of women before their first marriage.
The results are tabulated in Table 4.3. All groups now show a younger age at first
marriage for the renting sample, including white women for whom this pattern had been
reversed in the arithmetic mean calculation. The renting population married earlier than the
owning population by a margin of 2.35 years for black women, 3.04 years for white
women, 3.98 years for black men, and 2.13 years for white men. As suggested above,
this difference could be related to the fact that within the owning sample, considerations of

property transmission imposed greater constraints on marriage.

107

Table 4.3 Age at First Marriage, Vinton, 1910.

 

 

 

Group N Mean Age, First Singlulate Mean Age at
Mariage‘ First Marriage“

Black women, renting 319 19.7 19.25

household

Black women, owning 50 21 21.6

household

Whitewomemrenting 40 21.1 20.16

hwsehold

White women, owning 40 20.9 23.2

household

Black men, renting household 280 23.1 21.68

Black men, owning household 42 25.6 25.66

Whitemen,renting household 40 21.1 20.16

White men, owning household 38 25.6 22.29

 

" Arithmethic average
" CalculatedafterHajnal 1953.

 

108
5 li [El'l'l 1| II.

The fertility of late 19th and early 20th century Southern Americans, especially in
rural farming papulations, which are in many areas predominantly black, has received little
attention in recent years. Significantly more attention has been given to fertility in Europe
and less developed countries. General demographic characterizations have been presented
for the region by Eblen (1974), Engerman (1978), Coale and Rives (1973) and Farley
(1970). These scholars have documented that black and white fertility in the South was
generally higher than in other regions of the U.S.; and that both declined during this
period, white fertility first, followed by black fertility. Coale and Rives (1973) calculated
total fertility rates for Southern black women of 7.26 in 1880-84 and 4.84 in 19 10-14.
This suggests a decline of 33 percent. Similarly, Farley (1970) documented a decline in
the black child-women ratio of over 20 percent from 1880 to 1910.

Despite the severity of these declines, there is still little understanding of black
fertility during this period. The majority of work in Southern historical demography of the
postbellum period can be clmracterized as a debate over the uniqueness of the decline in
black fertility. This decline has been attributed to a variety of physiological factors such as
pellagra (Farley 1970), rickets (Cutright and Shorter 1979), venereal disease (Wright and
Pirie 1984), tuberculosis (McFalls and McFalls 1984), and general poor health. Other
researchers have emphasized the similarity between the decline of black fertility and the
earlier fertility decline among white Americans. These investigators have stressed
voluntary fertility limitation and looked to changes in the social environment and family
structure or economic value of children (Engerman 1978; Lantz and Hendrix 1978,
Masnick and McFalls 1976, 1978; McFalls and Masnick 1981; Meeker 1977; Thompson
1922). Until the recent work of Tolnay (1981, 1983, 1984, 1986, 1987, 1989), however,
almost all of this work has been hampered by a lack of sophisticated techniques with which

110

its decline several decades before black fertility, and to the extent that this decline is the
result of family limitation, or departure from natural fertility, a higher m value should be
obtained from the analysis. It is also expected that the Vinton area will show a higher level
of fertility, as has been demonstrated in many other studies of rural populations. Within
both the Vinton and West Point samples, the black subsamples should show higher fertility
than the white subsamples.

One of the strengths of this analysis is the ability to examine differences in fertility
according to tenure. It is hypothesized that there will be salient differences in the fertility of
owners and renters. However, because it is expected that the rural population was
essentially a natural fertility population, these differences will be expressed in differences in
the level of fertility (M and other measures) rather than its age-specific schedule (m).
Differences by tenure should be stronger within the urban sample, which is hypothesized to
show greater departure from natural fertility. However, it is also recognized that the tenure
variable is less structurally important in the urban environment, which may weaken
differences between renters and owners there. The next sections present the results of
several types of fertility analyses. Because the methods are complicated, with variable
parameters that are applied by the analyst, the manner in which these results were calculated
is laid out. The implications of the analysis are discussed in Chapter 6.

One of the basics of fertility analysis is the calculation of fertility rates. Usually
these rates come from vital registration data, such as birth registrations, and yield numbers
of births per number of women of a given age. However, since birth registration was not

instituted for the sample population in 1910, the best data available are census

lll

enumerations. This situation, which is not uncommon in many parts of the world today,
hm added impetus to the development of methods to convert census enumeration data into
rates. One such method is the own-children method, pioneered by Grabill and Cho (1965)
and Retherford and Cho (1978). Total fertility rates and marital fertility rates were
calculated for the sample population using the own-children method of estimation.

The crucial pieces of information needed to calculate fertility rates from census data
are a listing of a household giving the ages of children and the age of their mother. This
makes it possible to determine (at least within one year) the age of the mother at the time of
the child’s birth. Information is also needed on the number of children a woman has borne
and how many are surviving, to help in assessing the relationship between the children
listed in the household and the potential mothers. The fertility rate is calculated by first
deriving, for any given period of time, the number of children born to women at age X,
divided by the total number of women at age X. This procedure is carried out for all ages
of women between 15 and 49, and the final age-specific rates are summed for the total
fertility rate. For example, to arrive at the fertility rate for women age 30, the number of
children age 0-1 and whose mothers are age 30 is divided by the total number of women in
the sample population who are age 30. Then the number of children age 1-2 whose
mothers are age 31 (who would have been age 30 at the time of the child’s birth) is divided
by the total number of women age 31 . These two figures are added together, and the
procedure is continued with the number of children age 2-3 with mothers age 32, divided
by the total number of women age 32, the resulting quotient being added in to the
calculation. This prowdure is continued for a number of years, as determined by the
analyst.

This procedure could be carried out for a number of years before present, or ages of

children, as long as the data on ages of children and ages of their mothers were available.

112

However, the method is intended for use with census data, which shows only those
children present in a household with their mothers. As children are increasingly away from
their parents’ household as they age, the calculation should be kept to a low number of
years since birth (or age of children). The most common procedure is to calculate the rate
from one to fifteen years before present, or to age 15 of children. In this analysis,
calculations were carried back five years (or to the age of 5 years for the child). The result
is a fertility rate over five years, which is a reasonably short period of time. Calculations
over longer periods of time become increasingly problematic if there was much change
occurring in the fertility of the population, as is often the case. After the fertility rate for
age 30 is calculated, the procedure is continued for age 29, 31, 32, and so forth, until all
ages from 15 to 49 lave been covered.

Because the ages of children indicate their birth in terms of years before present, the
sum of the fertility rates where the child is 0-1, for all ages of women, can also be used as a
fertility rate for the presan The sum of the fertility rates where the child is 1-2, for all
ages of women, can be used as a fertility rate for one year ago, and the sum of the fertility
rates where the child is 2-3 can be used a fertility rate for two years ago, and so forth.
However, unless the sample is very large, it is more reliable to pool the data into a total

fertility rate for the total five year period.

Although this is the basic methodology for calculating a total fertility rate from own-
children information, a number of adjustments are also necessary. The own-children
method relies upon the presence of children with their mothers, so that the mother’s age at
birth can be estimated. Children not living with their mothers cannot be used. In the
sample population, relationship to head was listed, which generally simplified the task of
connecting children in households to their mothers. The fact that the census provided the

number of children a women had borne, the number which were surviving, the women’s

l l 3
marital status, and the number of years in a marriage, helped greatly in assessing the
relationship between a child and a woman. Although most children under 5 years of age
were living with their mothers, there were some exceptions. For example, some
households included nieces, nephews, or grandchildren of the head of the household.
These children could be included in the analysis only if the mother was also present, which
was not always the case. Also, children in simple nuclear family households whose
mother either was not present or whose age was not given could not be included in the

analysis. Fortunately this was rare.

A more frequent complication arose when a woman was married to a man who had
been married before, creating the possibility that some of the children in the household had
been home by an earlier wife. These children were still just listed as ”son” or ”daughter,”
and in fact many probably were the biological offspring of the male head of the house. To
handle this problem, the following procedure was used. If the number of children present
in the house matched the number of children the women reported as surviving, it was
asmmed that all of the children had been borne by the present wife. If the number of
children listed in the census exceeded the number the woman reported as surviving, the
ages of the children, the number of years the woman had been married, the birthplace of the
potential mother, and the birthplace of the child and its parents were used to make an

assessment.

For the entire sample, each child was assigned a ”status” number indicating the
surety of connection to a mother. These statuses are as follows: status 1) unknown (7), as
in cases where the mother was not present, her age was not given, or the household
structure indicated that this child was not the biological offspring of the mother. This was
the case, for example, where there were more children in the household than the woman

had listed as surviving, her husband had been married before, and this child was older than

114

the number of years she had been married; status 2) probable, applies to cases where the
husband had been married before but the number of children present in the household did
not exceed the number the woman listed as surviving, or the age of the children did not
exceed the number of years she had been married, status 3) good, indicates that none of the
above conditions apply. Within status 3, the woman could have been married before and
so the children might not all be from this marriage (not important in calculating the fertility
rate but of interest); and status 4) good, is used when the conditions of status 3 apply and
both parties have been married only once, so that all children are assumed to be not only the
offspring of this mother but this father as well.

Table 4.4 shows the breakdown of children into these statuses, by residence,
tenure, and race. The incidence of status 1, where the connection between the child and
mother was very questionable, made up from 0 to 12 percent of the sample of children,
depending upon the subsample. Table 4.4. also presents more information on the
circumstances of these ”status 1” children. Most common was that they were living with
grandparents, aunts and uncles, or just their father. Status 1 children were much more

common in black households.

In some subsamples as many as 15 percent of children could not be connected with
mothers with surety (i.e. the 1 and 2 status). This casts doubt on analyses which merely
assume that all children in a given households were borne to the resident adult female, at
least where extended families and more complex household structures are common, as in
the black sample. The incidence of children in households where mothers were not
present, or could not be confidently identified, increased dramatically with the age of the
child. Although this information is not presented in Table 4.4, 32 percent of black urban
children of all ages (< 17 years old) were status 1 or 2. The proportion of status 1 and 2

children was mallest for the urban white subsample (8 percent). Several factors may

115

 

 

 

 

 

 

 

 

 

 

 

 

2: 5e be ex we no to :2
oo— oo. 8. 9:30
R: we we we no 8 823 .83—
we 2: _e no 3 o gal
2 no 8 a 8 8 8 gel
8. we we we co no 8 833 a:
3 mm 2 me me we 2. 2 no 3 Eva”
m an no N... 8 mm 8 final
2 eo mm 1 no we we 2. 2 no co x a 1.5%
mm 2 mm 2 mm mm me mm 8 N— 88:”
cm 9. 9. 3 3 R on 3 8 g
R w_ wm a mm mm co mm we 2 335
med
:2: meeluém a 4.... a £5 E g
g twig. we: on. «o: 2: a a a .vlccd E M
a % win—fl g
o . a s as... as as" a: 5.. was as. is m w. N a sad
cote-emu new? .3 v9.58 Esp—ow $.88 _ 58w Bo E w v :o .0 3210508 Evade—05 we gm

Aegean—axe goes .3 5.6—Eu _ 258m He egg 2: Eat :0 8.5»: $38» :08 E 2e 88» n v 5.330 .«e $559K 8: co. co 8:55

.8.8 bane 2 e a BB .8238: a 8.229 a 3.5m .3. use.

 

116

contribute to this pattern. For a variety of reasons, children are more likely to move to
other households as they age. In addition, families become increasingly complex over the
life of the child, and this hinders the identification of a definite mother-child relationship.
Thus it is preferable to calculate the fertility rate over a relatively short period of time (or

ages of children), such as the five years used in the present study.

All children in status 1 were excluded from the total fertility rate calculations. To
adjust for the absence of these children, the number of children observed for each age
group was multiplied by a figure which essentially restores status 1 children and distributes
them in the same manner as those children who are with their own mothers. This
adjustment is the multiplicative inverse of the proportion of children in status 24,
ealculated separately for each age of child. The exact adjustment factors are given in Table
4.5. Separate adjusters were calculated for all households, and for only those households
in which the woman was married. The former was used in calculating the total fertility rate
and the latter was used in calculating the marital fertility rate. The incidence of children not
with own mothers was slightly higher in the total population than in the married population,

for all age groups but one.

Besides this adjustment, several other adjustments had to be made. Since the
method relies on counts of existing children, it does not include those children who did not
survive to be enumerated in the census. Thus the next adjustment is for the mortality of
children. This adjustment is made with reverse survivor methods based upon life tables.
Life tables calculate, among other things, the number of members of any one cohort (age
group) that will survive a certain period of time, usually a year, under given mortality
levels. By multiplying the observed numbers in each group by the multiplicative inverse
of the proportion of each age group that will survive a given year, deceased members are
added back into the fertility rate for each age group. Since males and females have different

117

uses u.2.... 3. a .25. 532 sass 2...: a. v. as. .83 .8... 9565 aa «.25 ea. 3.23.8 a...
2... cos... a... 280 as. Basses .2.
8... «Eon .5 “5.8 as. 335.8 ..

 

 

 

 

.3 8.. m... we. 2.. a... mum
.N. 3.. 8.. on. .2 3... al..
.N. 8.. N... 3... a... a... 3
.~.. 8.. 3.. 2.. an. 6.... WI.
.3 8.. 3.. 2.. an. 2.. 3
n... Mal...
8.. 3.. 8.. 2.. 8.. a... m...“
8.. . . .N. .N. m... m
8.. . . 8.. 8.. a... mm
8.. . . m... m... n... 3
8.. . . w... w... N... 3
u an».
:8 BE... 29:3 .52 as .3 8.5.2.. the... $8.55.
assess... 522.6 .56 6.2.6 .56 5.3.2.. 5.1%... £128. 2.5 a? 920

 

.o. a. as. base €220-58 £2229 2 9.55.8.6... .3. use.

 

118

mortality levels, separate indexes are calculated for each sex and then averaged using a

weight of the sex ratio to arrive at an overall adjustment.

The most important factor in this procedure is the availability of life tables which
approximate not only the overall level of mortality, but its distribution by age in the subject
population. In this case, model life tables provided by Coale and Demeny (1966) were
utilized. These tables have been calculated to simulate mortality by ages for different parts
of the world, based upon empirical evidence and mathematical modelling. The tables used
in this analysis were North Level 7 for the black population, and West Level 14 for the
white population. These are the same tables used in Tolnay’s (1981) analysis. The West
tables are designed for use in the United States (as well as other places), with Level 14 best
matching the known life expectancy rate of the time. A North rather than a West table was
used for the black population because it better incorporated the high incidence of
tuberculosis in Southern black populations (Tolnay 1981:445). The North Level 7 life
table best approximates the life expectancy and infant mortality for a turn of the century
black Southern population, as calculated by Eblen (1974). Table 4.5 shows the mortality
adjustments used in the analysis.

The final adjustment to the observed numbers of children by ages of their mothers
was for under or over counting by the census taker. Estimates of undercounting for this
time period have been made by Coale and Rives (1973) for the black population, and by
Coale and Zelnik (1963) for the white population. A multiplicative inverse of the
proportion of the population thought to have been missed was applied to the observed
numbers of own-children. This information was not available for specific ages between 0
and 5, but rather as one estimate for all children 0-5. As can be seen in Table 4.5, this
adjustment is a substantial one, especially for the black population. By applying this
adjustment to the entire number of observed children, it is again assumed that these

119

undercounted children are distributed among women of various ages in the same manner as

those children who were counted.

It was also newssary to adjust for mortality and undercounting of women used in
the analyses. These adjustment factors were calculated in the same manner and from the
same life tables or underenumeration estimates as described above for children, and were

applied by five year age groups. They appear in Table 4.6.

The resulting total fertility rates, and total marital fertility rates, are presented in
Table 4.7. Both rates are calculated in the manner described above except that the
population for the total marital fertility rate is limited to married women and children in their
households. The marital fertility rate is generally higher than the total fertility rate because
of the higher incidence of childbearing within marriage. Variation in fertility levels
according to racial and residential groupings is the same for both the total fertility and total
marital fertility rates. In both cases the lowest fertility is seen in the white urban
population, with the next lowest fertility in the black urban population. This is followed by
the black rural population, while the white rural population shows the highest fertility.

Q|l E IT! M

The total fertility rate discussed above is one of the most sophisticated fertility level
measures available for census data. It corrects for biases and provides a summary measure
which takes all age groups into account. There are, however, a number of other measures
which can also be utilized, some very easily. One such measure is the child-woman ratio,
which is the number of children under age 5 in relation to the number of women age 15-49.
Within this measure, more elaborate adjustments can also be made for the age-structure of

the population, but these were not undertaken in this analysis since the total fertility rate

120

Table 4.6. Adjustments to women, own-children fertility rates, 1910.

 

 

 

 

 

WHITE BLACK
N rmda'count‘ mortality "‘ N raidercormt” mortality‘"
1112 6 1.02 1.02 38 1.06 1.04
M 35 .99 1.03 117 .97 1.05
25:22 41 1.03 1.03 249 1.05 1.06
M 42 1.07 1.04 88 1.22 1.07
15:22 46 1.07 1.04 90 1.07 1.07
M 40 1.12 1.05 64 1.22 1.08
45:12 27 1.15 1.06 114 1.15 1.10

 

 

‘ calculated from Coale and Zelnik 1963

” calculated from Coale and Rives 1973
‘" calculated from Coale and Demeny 1966, whites, West level 14; blacks, North level 7

Table 4.7. Fertility measures, 1880 and 1910.

 

 

 

Group 1880, child/ l910,child/ 96 change 1910 1910 1910, 96 of
woman ratio woman ch/woman total marital woman married
ratio ratio, fertility fertility 10 yrs
1880-1910 rate rate thatarechildless
121m 460 370 -19 3.44 4.82 18.6
mm 710 620 -13 5.45 6.93 5.9
mm 650 320 -51 3.03 4.18 6.4
w 530 640 +20 5.67 7.30 5.8

 

 

121

was calculated. The child-woman ratio is presented for the sample population for both
1880 and 1910, in Table 4.7. It is presented by tenure groups in Table 4.8

It can be seen in Table 4.7 that the child-woman ratio shows the same ordering of
the racial and residential subsamples as the total fertility rate. White urban women have the
lowest fertility, followed by black urban, black rural, and white rural women. Also note
the sharp drop from 1880 to 1910 in the number of children per woman in all but the rural
white sample. The rural white sample for 1910 is rather small (only 101 rural white
households compared to 585 rural black households, 317 urban black households, and 201
urban white households), and the increase in fertility for the rural white sample may be
influenced by sampling bias. These changes between 1880 and 1910 represent declines in
fertility of 19 percent for the black urban population, 13 percent for the black rural
population, 51 percent for the white urban population, and an increase of 21 percent for the
white rural population.

Many other summary measures of fertility are possible. Table 4.7 also includes a
measure of childlessness, presented by racial and residential groupings. Those marriages
that were at least ten years old and in which the woman listed the number of children born
as 0 were tabulated, and are expressed in Table 4.7 as a percentage of all marriages of at
least ten years duration. The percentage of childless marriages was especially high for
black urban women, suggesting that childlessness could be a more important component in

black urban fertility rates than in other groups.

Several measures were not only calculated by residence and racial groupings, but
also by tenure. Unfortunately, total fertility rates and total marital fertility rates could not
be calculated by tenure because it was felt that the differences in standard of living

represented by these distinctions created different levels and schedules of mortality.

122

Furthermore, life table estimates were not available by tenure. However, several other
measures of fertility, which did not require adjustments by mortality, are presented in Table
4.8. The first measure presented is the child-woman ratio. This measure shows that
numbers of children per woman are consistently higher among renters than owners, for all
subsamples. The difference between the renters and owners is higher in the rural

subsamples, probably because the tenure distinction has more meaning in the rural setting.

The next measures presented are mean number of children born per year of
marriage, first for women who have been married less than 5 years, which is a measure of
early childbearing, and then for all married women, regardless of duration of marriage. It
should be noted that this analysis was restricted to women who had only been married
once, since those in multiple marriages may have had less exposure to pregnancy. Rural
whites are generally in the lower ranges in these measures, in contrast to their high fertility
as indicated in other measures presented above. The white samples usually show lower
numbers of children born than the black samples; and within each racial grouping, the
urban sample is usually lower than the rural sample. The exception to this pattern is the
higher mean number of children bom to urban whites, which is strongly influenced by the

high figure for urban white owners.

Figure 4.1 shows that these two measures follow basically the same pattern within
the white sample, with the lowest fertility indicated for urban renters, followed by the rural
owners, then rural renters, and with the highest levels among the urban owners. Like the
white sample, the black sample shows the lowest fertility among the urban renters, in both
measures. However, the two measures show different orderings from the second lowest
to highest positions. The rural black owners show an especially high initial fertility, but a
rather moderate level (compared to the other subgroups) when marriages of all duration are
considered.

123

mee- .mhe_ .whe_ Teach v5. 0.30 5.30:8 3.2.6.3 as

2.3... 9... =3.» 3...»... 8.8%... a

 

 

 

 

2.. een. e9. 2. 2. he be c... en. ww. .2.. maul
2. .3. 2.. 2. 2. as o... o.“ .n. 3. 3n. 915ml
2. 2. 2... 2.. an «n. 3. 3. a
3. mm. 3.. a: a... N5 N... o.w on. ow. «an. gum—I.
no. «we. 3.. a: .2. we «.0 n6 ee. 2.. man. mugs
.2. 2. mm... mm... 2.6 eN. mm. mm. .555...
cw. ~.. 9.. o.w a... e... c... e... on. 3. one. mafia
me. So. w.. a: o.w o... o.w 2. Nw. Ne. now. p.256
o... N. a. o. a... 3.. S. 3. 3.3.4..
no. 3.... 3.. a: 2. 0.0 o... od en. no. wen. mag
eoo. Nee. ea. .2. ed aw a... o.w «w. ee. awn. as
2. ed :6 n2.“ o... .w. 3. um. 53...
G. I O I O I O I
3.6.8: wmém one wmém owe wném one 2.3% one 3.3 one he... .1 in v5... 2....
8.30m . s. o s m 03.52.. N 93.5.... .ouaEa... veg... =¢ .2623 .5... S 5...... .a case?
:8... a 2 ER. 5 .. ER. 5 . ES 5 s 58 .e .. ER. 6 a .8 3.5 .8 2.5 :55 .85

 

 

.2 e. .838... .050 E... newsman... be 5.35.. .3 Eon 5.6—Eu he E3552 .w.w 055.

 

124

 

.omm._tmE .0 Ln?» .8 Eon c8220 .6 39.5.. @0803. .3» 959;

sumac; 3.30:; Lamas; .5302; 3.3.; 3:505 .53.; .5305

 

 

 

 

 

 

 

 

 

 

 

 

 

22.88 23.20.: :0 Le 88 8.9% 682%.: B 28> m2. .8: Le 28 2.8

0.0
_.0
N6
m.o
v.0
m.o
0.0
No
w.o
0.0
o.—

125

This pattern could result from initially higher fertility for black rural owners,
followed by lower fertility than other groups. One might expect this pattern in conditions
of parity-dependent target fertility where conscious efforts are made to control fertility
according to numbers of children already born. However, other fertility measures
discussed below do not support this interpretation. Among all but urban owners, the black
sample has a higher rate of early childbearing than the white sample. The black sample
actually shows much less variation in numbers of children born by residence or tenure,
especially when all durations of marriage are considered, compared to the white sample.
Because of the influence of age on fertility, the measure for the first five years of marriage,

which should tap a more unifomly younger sample, is probably more accurate.

In contrast to the above measures, which look at the intensity of childbearing,
another summary measure of fertility is completed fertility, the number of children born to
women who have completed childbearing. This measure is usually computed using the
birth histories of women of various ages over 50 years old. Because of the small sample
size at 55+ ages, I use the 50-54 year age group. Some additional births could occur after
age 54 but should not be significant. The number of children born to women between 50
and 54 are tabulated in Table 4.8, for various marital statuses, and by tenure, race, and
residential groups. The comparisons by different marital statuses show the lower fertility
of widows, although the difference is slight for the black sample. The effects of multiple
marriages on fertility can be seen by comparing the mean number of children born to
woman of different marital statuses. There is no uniform pattern between the number of
children born to black women in a first versus a second marriage. Black women in third
marriages do show significantly lower numbers of children born than those in first or
second marriages.

126

Figure 4.2 shows completed fertility. On the one hand it can be observed that the
patterns by residence and tenure do not follow the same path for the black versus white
samples. In the black sample it is the rural renters who have the highest fertility, while in
the white sample rural renters have the second to lowest fertility. The highest fertility in
the white sample is that of the rural owners, who have the lowest fertility of the black
sample. Two patterns are visible. One is the tendency, within any racial and residential
grouping, for owners to have lower fertility than renters. This holds for all groups except
the white rural sample, which is the smallest of any group. The other pattern is that within
all residential and tenure groupings except rural renters, the white sample has higher
completed fertility than the black sample.

The final summary measure of fertility levels to be discussed is the measure M. M
is the ratio of the age-specific fertility of the study population to a series of natural fertility
populations (Coale and Trussell 1974, 1975, 1978). For example, an M of 1.5 indicates
that the fertility of the study population is one and a half times higher than the fertility of the
series of natural fertility populations studied by Coale and Trussell. This measure has
become a standard one since its adoption by the Princeton University European Fertility
Project. M is calculated along with another measure, m, which assesses the departure of
the study population from the age-specific fertility schedule of a natural fertility population,
and is useful for estimating family limitation. Since 111 will be discussed in more detail
below, the methodology behind the calculation of M will be deferred until that time.
However, the findings from the calculation of M will be noted here. M is tabulated in
Table 4.8, where it can be seen that the fertility levels of the sample population are usually
high, often more than one and a half times that of the model natural fertility populations.
Of most interest is the ordering of the various subsamples of our population. As with the
child-woman ratio and total fertility rate, M indicates the lowest fertility among urban
whites, followed by urban blacks, and rural blacks. The highest fertility is seen among

127

.20 88> vmém coEoB 87:9: 8 Eon cons—Eu .o a ommno>< NV 83:

guano
3%ch 3.26:2, .5322, 526:2, 3.an 3.265 .5an .5305

 

 

 
  

. -

 

 

 

 

 

 

 

 

 

 

 

 

 

128

rural whites. Within the racial and residential groupings, renters have the highest levels of
fertility, except within the urban black subsample.

Table 4.9 presents a summary of the ordering of fertility for the various racial,
residential, and tenure subsamples, according to the different fertility measures discussed
so far. The considerable variation between different fertility measures is evident in Table
4.9, especially when the samples are broken down into tenure groups. Some of the
fluctuations may be due to the small size of the rural white sample, which behaves
erratically. For example, while the rural white sample displays the highest level of fertility
in several measures, in other measures, such as children born per year of marriage, it is
low. Of course, the difference in these two measures is not necessarily incompatible.
Because of increased exposure through longer duration of marriage, the rural white
subsample may have a higher overall fertility, in spite of lower fertility per year of
marriage. The most reliable measures, such as the total fertility rate, M, or even the child-
woman ratio, consistently suggest that the lowest level of fertility was in the urban white
sample, followed by the urban black sample, the rural black sample, and the rural white
sample. A most interesting observation is that there is more similarity in the overall rates
by residence than by racial groupings. That is, for example, the urban black rates are
usually more similar to the urban white rates than they are to the rural black rates. The
implications of this frnding are discussed below. The effects of ownership or tenure are
less comistent, and it is more difficult to generalize since a total fertility rate could not be
constructed by tenure. However, using the child-woman ratio and M, probably the two
most reliable measures of those broken down by tenure, renters showed higher fertility

than owners in seven out of eight comparisons.

129

can. .3 58.85 8.... Bases .83.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

w n m . made
5 m N o m.0..3°
:3 .83— .. v v v m m _ :3 .83—
m n . m «.02.».
. v o _ €0.50
=3 :35 . . . . n m N 3
o o n N. 20...».
m _ w n €0.56
5 3.3— m m m n N v 5 m
N n v v E059.
v N n N 80:30
3 :35 N N N N _ m .— :35
E. v as. .69. 2.. n v
Vmén one .20 Eu... 0:... ...0.._ 03% oz... one 5.. En...
Eon Saab.— ..553 .. are: .3520...— ....:.93 E3 80mg
2 5.220 # 5.6—Eu >220 2 30,—. .58. \220 £5: 5.230

 

3.55. as»... a E .8.8. as. .2385 as a «5.88. 2.5858 .08 .o .5. a...» use.

.2 a. £0.83... 3.5.0.. 0455...... meatanU .eé 033.

 

130
II 5 _5 'fi Ell [E I'l'!

Although some of the above measures were calculated by ages, they were basically
designed to esfimate the overall level of fertility in a population. However, the way that
fertility varies by age is important, especially when considering the degree to which fertility
might be subject to deliberate control or intervention. Fertility is known to fluctuate with
age because of biological variation or fecundability, but in a regular manner. This
variability has been modeled, so that populations which depart from this pattern can be
picked out. These populations are then assumed to be deliberately controlling fertility
(Coale 1971; Coale and Trussell 1974, 1978; Henri 1961; Knodel 1977; 1978). This
does not mean that only those populations with a specific age-schedule are attempting to
control fertility, since deliberate efforts to space children can result in an age-specific
fertility pattern similar to natural fertility (Knodel 1987). Historical demography,
especially in the West, has shown that most fertility change over the past several hundred
years has occurred via a change in the age-specific schedule of fertility behavior, especially
through increased reduction in fertility at the later childbearing ages and higher parities.

The measure m indicates the degree to which a population departs from the age-
specific fertility schedule of a group often natural fertility populations, as compiled by
Henri (1961), Coale (1971) and Coale and Trussell (1974, 1975, 1978). The specific
methodology for calculating m has been supplied by Coale and Trussell. It is calculated by
first constnrcting age-specific marital fertility rates for the population in general. Own-
children methods as described above were again used, and follow very closely the
methodology described above. However, one departure should be mentioned. In order
to conduct this analysis by tenure, the various adjustments used in the calculation of the
total fertility rate were not applied, since adjustments for mortality and undercounting were

not available by tenure.

131

Table 4.10 shows the percentage of children born that were surviving for women
over age 50. There is considerable fluctuation by residence and racial groupings, as well as
by tenure. The survival rate was higher for owners for the rural samples, and higher for
renters for the urban samples. Since different adjustments were not available by residential
or tenure groupings, the data were accepted at face value for constnrction of M and m.

This is likely to create some error in the estimation of M, the level of fertility, probably
causing M to be a bit low. This error, however, should not have a large effect of the
measure m. As a test, the adjustment factors were added to one set of calculations of m
and M. The change in m was less than .01, and the resulting mean square errorwas
increased, which suggested a better fit of the raw data to the model fertility age-schedule
without the adjustment.

Table 4.10. Children born and surviving to women age 51+.*

white renters white owners black renters black owners

 

Vinton (N) (3) (6) (59) (16)
# born 6.0 6.33 7.38 5.87
# surviving 4.33 5.33 4.55 3.6
96 surviving 75 88 63 71
West Point (N) (12) (24) (18) (33)
# born 6.75 6.79 7.83 5.6
# surviving 6.09 4.66 3.72 2.9
96 surviving 87 73 52 48

‘all figures except Ns are means

 

Once the age-specific marital fertility rates were computed, they were compared to
the model schedule of ”natural fertility” rates, and least squares regression was used to
estimate M, the level of fertility, and m, the degree of departure from the natural fertility
age-schedule. Coale and Trussell (1978:203) state that

132

”in any population, the ratio of marital fertility r(a) to natural fertility n(a) at
a specified age is given by

r(a)/n(a) - M exp [m " V(a)1.”

They also state that ”the function v(a) expresses the tendency for older

women in populations practicing contraception or abortion to effect

partiarlarly hrge reductions of fertility below the natural level” (Coale and
Trussell 1974: 188).

The values of n(a) and v(a) as given by Coale and Trussell (1978:205) and as used
in this analysis are as follows:

Table 4.11. Values of n(a) and v(a) used for calculation of m and M.

 

 

 

Ame M9 30-34 35-39 40-44
n(a) .460 .431 .395 .322 .167
V(&) 0.0 -.279 -.667 - 1.042 -1 .414

 

Values for ages 15-19 and 45-49 were not used, as suggested by Coale and
Trussell, because of the high degree of fluctuation at the ends of the reproductive period.
The above eqration expressing the relationships between the M, m, r(a), n(a), and v(a) can

be rewritten as

lnlt(a)’n(a)l - lnM - m’V(a)

133

which can be reinterpreted in the form of

y=c+m(x)

Since r(a) is the age-specific marital fertility rate calculated for the population in question,
and v(a) and n(a) are supplied from Table 4.11, the M and m in the emation can then be
estimated using least squares regression. There are other methods to solve the equation,
but least smares regression is recommended by Coale and Trussell (1978) as giving the
most equal weight to every age group, and was the procedure followed in this analysis.

The results of this analysis are indicated in Table 4.8. Values of m range from
0.017 to 0.992. Coale and Trussell suggest that values of greater than 0.2 to 0.3 begin to
suggest fertility control, with modern contraceptive populations often having m values well
over 1.0. Although there is considerable ambiguity in values around 0.2 to 0.3, most of
the values calculated for the sample population are sufficiently above 0.2 to suggest some
deviation from the natural fertility schedule. Only the rural black population exhibits an
age-specific fertility schedule clearly similar to that of the natural fertility models. The
highest degree of deviation, and thus the best indication of family limitation, is in the urban
black sample, and within it, especially for owners. The urban white and the urban black
samples also exhibit an m value over 0.3, again especially for owners. For all three of
these subsamples, the m value for owners is considerably higher than that of renters.

The final column in Table 4.8 gives the mean square residual, which is an
indication of the degree of fit of the solution to the equation. It can be seen that this
measure is fairly high, which indicates that most of the observed schedules of age-specific
fertility rates do not easily fit the model schedule. In fact, Coale and Trussell (1978)

134

suggest that mean square errors of more than 0.01 indicate divergence. The high degree of
divergence may partly stem from small sample size. It can only be hoped that the overall
finding in the degree of family limitation between the various subsamples is still accurate.
However, it is suggested that little confidence be given to small differences in m values,

because of this lack of fit.

While the overall summary m, which is presented in Table 4.8, is of greatest
importance, differences in individual age-specific fertility rates can also be informative.
Comparisons of single age-specific rates, however, are especially problematic for a sample
as small as this one, which can create large fluctuations in age-specific rates. These
individual age-specific fertility rates are shown only for the larger black subsample, in
Figure 4.3. The age-specific fertility rates [f(a)] are plotted against the age group for
which they correspond (only the beginning of the age group is printed on Figure 4.3: 20
indicating 20-24, 25 indicating 25-29, etc.). This technique is useful because the resulting
curve generally takes on a concave, as opposed to convex, look when there is a significant
degree of family limitation. This is because the sharper reduction in fertility that occurs in
the middle to late age groups under family limitation causes the curve to buckle downward.
Under a regime of natural fertility, the reduction in fertility over the ages is due to reduced

fecundability and is more gradual, resulting in a convex curve.

Looking at Figure 4.3, it is possible to compare the curves for the rural
subsamples, which exhibit the lowest m values, versus the urban subsamples, which
exhibit moderately high m values. Although the curves are interrupted by the erratic
fluctuations between the age 25-29 and age 30-34 age groups for the rural owners, there
are observable differences between the curves of the rural versus the urban subsamples.
The curves begin to buckle much earlier for the urban subsamples, at age 25-29 for urban

renters and at age 30-35 for urban owners, compared to the rural samples. This suggests

135

.5305 .0.

3.39m ..o.

.mEEmm .690 .99 .938 3:55. 2:880? .me 059“.

mom
0.. mm on mN 0N

 

 

.. m6
0.0

 

 

No

5&5 -_.. + no
so

Damm—m 1.. e O.—

.3.

136

more and longer term fertility limitation for the urban samples, to the extent that younger

age groups are also affected

m 'I'

The final aspect of fertility to be examined is whether births take place within
marriage. Because of the close association of conjugality and residence in Western
thought, illegitimacy is often associated with departure from a standard of household
composition and residence (the two parent nuclear household) as well as departure from the
standard of birth within marriage. Therefore, both the analysis of marriage and the analysis
of household composition and structure provide useful perspectives in this study.

In regard to marital status, it was shown above that marriage was a very common
experience for Vinton residents, being universal for black women, and almost as common
for white women. However, at the same time, the analyses showed that women were often
married more than once. Periods between marriages increase the possibility of illegitimate
births. The above analysis of fertility suggests that childbearing in the Vinton community
was not as broken out or as subject to manipulation as in later times (i.e. was more
”natural”). Thus the most important ingredient in the relationship between birth and
marriage is the incidence and nature of marriage. Factors such as knowledge or use of
contraception, and risk-taking strategies or behaviors, are more applicable in studies of
contemporary illegitimacy. They have much less impact here, heightening the relative
importance of a general understanding of marriage, marital status, and household

composition.

To understand illegitimacy in this population, a group of women who had

illegitimate births are identified and scrutinized. Direct data on illegitimacy is not provided

137

by the census, although careful use of the available data has allowed indirect estimation.
While the absence of direct data excludes speaking of either illegitimacy rates or ratios, this
indirect estimation allows an in-depth, individual level look at a number of women who had
an illegitimate birth around the turn of the 20th century in Vinton. Although men as well
as women can have illegitimate children, only women give birth and only women’s birth
history is recorded in the census. Thus, it is both theoretically and practically expedient to
orient this research to the life experieness of women.

Since this data on illegitimacy has had to be estimated from the household
enumeration, a brief explanation of methodology follows. The basic technique was to
compare the number of years a woman had been married to the age of her first born child.
Although some enumerators did provide information down to the month for marriages and
births, the more common pattern of rounding to a year meant that at least one year’s leeway
had to be given. For a women to be included in the sample, a series of specific types of
information needed to be present and a series of conditions met. These conditions were: 1)
a woman had to be in her first marriage, since the number of years married given in the
census referred only to the current marriage; 2) both marital status and number of years
married had to be provided; 3) both the number of children born and number of children
surviving had to be given and all children had to be surviving; 4) besides surviving, all
children had to be physically present in the household, with their ages given. Given the
presence of all of this information, the relationship of the first birth to marital status at the
time of that birth could be assessed. For example, a woman who had been married five
years, in her first marriage, and who had borne six children, all six of whom were
surviving and present in the household, and the age of the eldest being seven years, had
apparently had an illegitimate birth. Should any part of this information be missing, the
assessment could not have been made. In addition to these first married women, women

who had borne children but who had never been married were also included in the sample.

138

The strength ofthese restrictions means that only a small sample ofthe total
community, and probably only a sample of illegitimate births, was included. It may be
useful to reflect on the possible biases of the sample before proweding. Missing
information was generally not a problem. However, a major selective bias for inclusion in
the sample was total survivorship of children. Many women had some children who had
not survived, making it impossible to judge whether the oldest child listed in the census
was really the first born and thus how many years had passed since the first birth. This
perhaps biased the sample toward families in which children were physically stronger, or in
which better socio-economic conditions contributed to greater survivorship. Also, even
when all children were surviving, in many cases all of these children were no longer living
in the parental household. This also made it impossible to judge the number of years since
the first birth. This especially excluded older women from the sample, since their children
would be older and less likely to be living at home. While this bias is not especially
alarming, it is important to note that the birth and marriage experiences discussed are
predominantly those occurring over the past 15 years, from about 1895 to 1910, during
which time these women would have been approximately 15 to 30 years old. Birth

experiences occurring much earlier are less represented by the sample.

One hundred nine women met the strict criteria set out above. They form a lowest
possible estimate of women having had an illegitimate birth, since so many women could
not be considered due to the absence of some children from their households. These 109
women represent 6.8 percent of the total sample of 1,601 women 16 years or older. The
average age of the 109 women was 29 years, as opposed to an average age of 36 years for
the total 1,601 women.

139

Table 4.12 presents the sample group by ethnicity and tenure of farm or house.
The most obvious result from Table 4.12 is that illegitimate births, at least as gathered from
data reported to the census taker, occur more frequently in the black sample, and more
frequently in the rural sample, for both blacks and whites. The largest category of women
having had an illegitimate birth is the rural women from households that rent their farms
(N- 68, the next highest category being 19). These 68 women make up 11.8 percent of all
women in black rural households that rent their farms. A higher percentage of women with
an illegitimate birth occurs in the sample of women from nrral black households that own

their farms, where at least 16.9 percent of women had borne an illegitimate child.

Table 4.12. Women with illegitimate births.

 

 

Blacks Whites

N Percent of all black women N Percent of all white women
Vinton sample
renters 68 1 1.8 4 6.9
owners 1 1 16.9 1 2. 5
West Point sample
renters 19 8.0 0 0
owners 1 l 5. 3 1 .4

 

No clear pattern emerges as to any relationship between the proportions of women
having had an illegitimate birth and land tenure. Opposite patterns are observed for the
rural versus urban samples -- lower illegitimacy for niral black renters compared to rural
black owners and higher illegitimacy for urban black renters compared to urban black
owners. This opposition suggests that different relationships between tenure and marriage
may operate in the rural versus urban setting. The white sample is so small that
comparisons by these detailed breakdowns seem unsupportable. However, it should be
noted tint the highest proportion of white women having an illegitimate birth is for the rural

140

rent farm category, and that the percentage here, 6.9 percent, is only slightly higher than
the lowest percentage for any of the black subsamples.

The higher proportions of woman having had an illegitimate birth are in the rural
setting, despite the fact that sex ratios are 90 (males per 100 females) for the rural sample
and 75 for the urban sample. The very low sex ratio for the urban sample is most likely the
result of extensive rural-urban migration of black women, who could easily find work in
town, usually as cooks, laundresses, or general domestics. There was generally less work
for black men in small urban centers at this time.

Relatively higher illegitimacy in the rural sample can also be seen in another
measure, that of the number of years between the illegitimate birth and subsequent
marriage. This measure was calculated by visual inspection of each case. For the rural
sample, the maximum number of years between the illegitimate birth and marriage was 11
years, with a mean of 3.3 years between the birth and marriage. In the urban sample, the
maximum number of years between the illegitimate birth and subsequent marriage was 9
years, with a mean difference of 2 years. Again, this is despite the fact that the lower sex

ratio for the urban sample would suggest that urban women might lmve greater difficulty

finding a spouse.

With these general observations on the tenure and residence of women with
illegitimate births, it is now time to take a closer look at the marital and residential status of
the group of women who had an illegitimate birth. Of the 109 women, 14, or 12.8 percent
were single. The rest were in their first marriage (the only other choice given the
restrictions of the analysis). Of these married women, only two out of 94 women had
children with more than one surname or with a surname different from their own and that

of their spouse. That is, in the majority of cases, the illegitimate child bore the name of the

141

subsequent spouse and legitimate children. The census enumerators sometimes noted if
children were adopted; in none of these cases was there any such notation. It is not
possible to say from these data if the arrrent spouse was the biological father of the
illegitimate child(ren), although it is assumed that the residence patterns (and stated
relationship of ”child” by the census taker) indicated social parentage.

What about the 14 women who had children and were not married? First of all, it is
important to note that they are all black. No white woman indicated that she was single and
had borne children. Their ages ranged from 18 to 45. Most had only one or two children,
although one single women age 35 had borne six children and another single woman age
35 had borne three children. Eight lived in the rural community and six in town. Of those
living in town, two were cooks, two were laundresses, and two did not list an occupation.
Of special interest is the residential status of these women. Being a single mother did not
necessarily mean these women lived alone with their children. In fact, only four of the 14
headed their own households and lived alone with their children. Seven, or 50 percent,
lived with one or both of their own parent(s). This residence pattern was not restricted to
the youngest women in the sample, and included one 35 year old woman and one 45 year

old woman. The remaining three single women lived in the household of a sister.

One cannot say from census data, or any other information on residence, or even
marriage after an illegitimate birth, how these births were viewed. But the fact that so
many of these women had gone on to marry suggests that the illegitimate birth did not set
them apart or create a stigma strong enough to rule out marriage. Although the mean
difference between years married and birth of first child was two years for the urban
sample and three and one third years for the rural sample, 67 percent of the women had
married within one year. It is possible that the birth, if not the conception, of a child may
have spurred marriage.

142

Most of these women only had one illegitimate birth. Of the single women, only
two women had more than one child. Of the married women, one woman had borne three
children before marrying, and three women had borne two children before marrying. All
others had borne only one child before their marriage. These figures, along with the other
data presented above, suggest the overall prominence of marriage, for both black and white
women. However, the higher incidence of illegitimacy for black women may stem from
subtle variations in the timing and duration of marriage, as well as the relationship between
marriage and childbearing.

CHAPTER 5
AGRICULTURAL PRODUCTION, TENANCY, AND THE HOUSEHOLD

Muslim

Chapter 2 presented a description of the basic nature of tenant farming in the
postbellum South, and gave a detailed background on the study area. This chapter lays out
the analysis of agricultural production, and relates it to the developmental cycle of the
household. The majority of the analysis is based upon 1880 agricultural census data,
which was linked to the 1880 household data, as described in Chapter 1. Use is also made
of a number of chattel deeds, or liens on crops, recorded in the Clay County chattel deed
books for 1879. The 1879 deeds correspond best with the data listed in the 1880 census.
Agricultural data for 1910 comes partly from the population schedule of the census, which
indicated whether a household owned or rented their farm. The population schedule of the
1910 census did not make a distinction between sharecroppers and renters, as was made
for the 1880 sample in the agricultural schedule. Analysis is also conducted utilizing a
ledger from the Vinton landlord, Henry D. Watson, (Watson collection n.d.), whose

operation was discussed more generally in Chapter 2.

As mentioned in Chapter 2, the study community was overwhelmingly rural, and
involved in tenant farming. Eighty one percent of households in the 1880 Vinton data set
were black. Of these black households, 6 percent owned their farm, 47 percent rented
their farm, and 47 percent sharecropped. Of the white sample, 64 percent owned their
farm, 25 percent rented their farm, and 11 percent sharecropped. In 1910, the percentage

143

144

of households that were black had increased from 81 percent to 90 percent. The percent of
the black household that were tenants had decreased from 94 percent to 85 percent. The
percentage of white households that rented had increased from the 36 percent of 1880 to 50

percentin 1910.

As was common in this black prairie region, agriculture during this period was
focused on cotton production. Corn was also a major crop and some vegetables, especially
sweet potatoes, were grown. The most important livestock were mules or horses, less
frequently oxen, for traction. Swine and poultry were most important for subsistence.
Most farms were arltivated with household labor, with a small percentage using part-time
hired labor.

Table 5.1 presents a basic description of agricultural production in 1880, broken
down by tenure and racial groupings. One of the most consistent patterns is that within any
tenure grouping, white farms usually show a higher mean level of production or a higher
value of basic inputs (machinery, livestock, or value of farm) than black farms. Also,
within either racial grouping, owners usually show the highest level of production and
wealth, followed by renters, and then sharecroppers. Although these relationships are not
without exception, they are overwhelmingly consistent. In many instances, the
differences between the black and white samples is so great that the poorest white sub-
sample, the white sharecroppers, have farms that are better supplied than not only black
sharecroppers, as would be expected, but also black renters. In fact, in 12 of the 20
columns, this is the case. Those cases in which the white sharecroppers show a lower
value or level of production than black renters are Columns 4, cotton bales per acre;

Column 5, cotton bales per worker; Column 9, money spent on wages; Column 12,

145

 

 

 

a: S 3. 3. NnN 9N mm. n2 ca. wow 2 5:3
a: 2. Nn mNN one. 3... mm. EN «N and cN 0.;3
C weN EN wa . eNNv N.o_ mm. new o3. N3— 3 30:3
m mm N 8. wow m.m 3”. mm. 2: omn Ne. 55
o. 5 on of Se _.m v. on. on. 3% e3 0.3
a: eN_ cw 3N omfl nod eN. 3N nmm mNS NN 3o...
- .33

ad a. a a a. a 3 a a a

bum—$0.. mowe3 2.2.2er :09. .8.... 3\ 0.08 0\ 3\ v3.60...
.5 .5 ho he be 8.3 8.3 we... so... 8V

2.09. m .5ng 02.3 m 02.3 m 02.3 ml .558 .528 2E3 0E3 033

 

6.5.0... 3 380.30 63. .m0...meo... 3.3—53.3.. ._ .n 03E.

 

146

 

 

 

.. N. .N on «.mN .... .... 8. an. ... WI. .23

.. v. .N e N...N .... N... nN. ... ...N IgloN ..3

.N 3. cm. 2N 0.8 N... 0.: .... ... .... l3 32.3

.. 2 mm .0 0.2 m... 3. 3. NW. .... Nlla. a...

a N. .N n. o... «.0 o... ... S. 0.. l... o....

NN 2 N. E 3: .2 .... 8. .... N... NIIIIN 3a...

mango
SN. 8.. a: a: G: a: ...: .N: .N: ....
EU“ $0.—UM $0.—U“ 8&0“ —>’O.u 0=m>¢m @360 §\ E 8:05.:

5:00 53:0 v0.9.2.5 33.0.5.5 5:00... .5908.

 

$3.80. ..n 030%

 

147

traction per worker; Column 13, traction per acre; Column 16, fowl; and Column 20;
cotton acres per worker. Several of these measures deal with productivity, suggesting that
while most indicators suggest wealthier and more productive farms among whites, the
productivity levels of white sharecroppers was not that different from that of black
sharecroppers. Sharecroppers would have the least amount of control over the inputs to
the crop,and would have the highest level of these inputs supplied from outside their own
household, which may explain the lack of differentiation between white and black

sharecroppers.

Traction is a very important variable in tenant farming. It was defined in this
analysis as mules, horses, or oxen, with cattle listed separately. It is shown in Columns
ll, 12, and 13 in several different ways. The higher traction of renters compared to
sharecroppers follows from theoretical understandings of these two arrangements (see
Chapter 2). The relationship between renters and owners in reversed in Column 13
because the acreage difference between owners and renters is proportionately greater than
the traction difference between these groups. The other livestock categories consistently
show higher numbers for owners. This would give owners a reduced dependence on
purchased food. The relationship between renters and sharecroppers regarding other

livestock is less consistent.

Besides traction, the other crucial variable in agriwltural production is acreage.
Here, owners show more cultivated acres, both absolutely, and per household worker
(defined following the age criteria discussed in Chapters 3 and 4 and shown in the
Appendix). Owners also have a much higher number of unimproved acreage.
Unimproved acreage was important in raising livestock and in hunting, again suggesting
that owners could have depended less on purchased food. Their higher number of

unimproved acres also suggests that owners could expand agricultural production more

148

easily than renters or sharecroppers. Chayanov’s work on the relationship between
agricultural production and demographic variables (1966) was conducted under the
assumptions that households had the opportunity to expand production if they desired.

Most of the productivity measures, such as Column 2, value produmd per worker;
Column 3, value produced per consumer; Column 4, cotton bales per acre; or Column 5,
cotton bales per worker, show considerably higher productivity for owners. Value
produced is a category provided by the census enumerator, and it is intended to estimate
both cash crops and the value of goods consumed at home. This difference is most likely
related to owners’ higher traction, better or more tools and machinery (Column 8), and
perhaps ability to choose the more fertile land, since their farms were larger. Higher
productivity may have been possible for owners because their higher level of wealth and
greater autonomy meant that they could hire extra labor at crucial points in the crop cycle.
Cotton is a very labor intensive crop, but one in which labor needs fluctuate dramatically
over the growing season. The harvest can be ruined fairly easily if it is not gathered
quickly, especially in a rainy season. Those farmers able to hire or otherwise mobilize
labor for that task would be at a tremendous advantage. However, it is interesting to note
that Column 4, cotton bales per acre (this is acres planted in cotton only) does not show the
usual pattern of higher yields for owners, followed by renters and sharecroppers. The low
figure for black renters is especially interesting, and may represent a decreased devotion to
the cotton crop by black owners, who would have had more autonomy, and perhaps

. lowered demands to focus work on cotton, than black sharecroppers.

This suggestion is not especially supported when cotton production is viewed
another way, however. Table 5.2 shows a number of variables that were constructed to
explore crop mix and agricultural strategies. Column 4, percent of cultivated acreage in

cotton, does not suggest that black owners were de-emphasizing cotton. Instead they

149

show the highest percentage of cultivated acres that are planted in cotton of any group, with
no clear difference between renters and sharecr0ppers. There is no overlap between the
black and white samples for the percentage of acres in cotton; all black tenure subsamples
planted a larger percentage of their land in cotton than white subsamples. This is probably
because of fairly similar needs for cash or rent from cotton, despite differences in size of
holdings. Corresponding to this is the higher percentage of cultivated acreage in food
crops for the white sample (Column 5). It should be noted that Columns 4 and 5 are not
inverse by definition, since other crops, such as hay, cats, or tobacco, were not included in
either calculation. Comparing Columns 4 and 5, the reader will note that they often sum to
over 1.00. This is undoubtedly because of estimation errors from the census, which
calculated total cultivated acres as a separate question from questions about specific crop
acreages. This category is not calculated by adding up the acreages reported for various

crops.

Several other variables provide an indication of differences between farms. Owners
produced more food, both absolutely (Column 6), and relative to the number of consumers
(Column 7), than renters, who were followed by sharecroppers. The pattern noted above,
that within any tenure group, whites produced more than blacks, is also noticeable, as is
the fact that white sharecroppers once again fall equal to or above not only black
sharecroppers but also black renters.

Two specific food crops investigated were com (Column 2) and sweet potatoes
(Column 3). Both were important dietary staples, although com was also used as livestock
feed. Both show higher levels for owners. Tenure seemed to have little effect on corn
production, and little effect on sweet potato production for the white sample. Black
sharecroppers showed lower sweet potato production than renters or owners. Similarly,
the black sharecroppers show an especially low ratio of bushels of food per bale of cotton

150

 

slams

 

 

n... o. N.. ..N. ... ... s ..
.... n. a. 3.. .... N.. v a 0.31....
n. .N ... .... 8. NN. .. z. a
...N n e. NN. S. 3. e . ... M...
Na ... on ... 3. N.. v . glen...
Na 2 3. an. ... 2. N .N 3

=20

.... E .8 5 .... 5 .N. E
.5300 3.3 0\ .50.. .08.. ... ......00 E 03.0.. €0.00 €0.00

38...»..3. ..8. 225...

mmhmam v.20... .x. 00.0w 8

.2. .0030. E8 ..o..8

 

.050... mu v8.8.9.0 63. 603—8600.. ...... ...... no.0 .8580... _n.:._3tm< N.n 030,—.

 

15 1
produced (Column 8). It is likely that this category is influenced by the increased control
landowners had over sharecroppers. The very high number of bushels of food for white

sharecroppers is puzzling, especially since they do not show a high percentage of acres in
food (Column 5).

 

The previous discussion has focussed on variation in agricultural holdings and
production by racial and tenure groupings. Because these social divisions show
importance, these divisions are retained while further examining variation by household

composition and life cycle stage.

Clmpter 3 showed that the majority of household in the study area were nuclear
family households composed of a married couple with or without children. Some,
although not many, households also included relatives of the head (or spouse), boarders,
and servants. Table 5.3 shows 10 summary agricultural measures according to household
type. Household types are the same as those discussed in Chapters 3 and 4: 1-2 is nuclear;
3 is single parent; 4-5 is adult siblings or other kin; 6 is unrelated persons; and 7 is solitary.
The first two rows for each household type present information broken down only by race,
with subsequent rows broken down by race and tenure. Because the subdivisions are so
numerous, and some Ns subsequently small, patterns are easiest to discern when looking
first at the divisions by race alone. Any given pattern is not totally consistent across all of
the variables (columns), or for all subsample within each racial grouping, but general
patterns can be picked out.

Initially it can be observed that among the black sample, household type 3, single

parent farms, seem to have less equipment, livestock or other resources than type l-2, two-

152

parent farms. This is the case for some but not all variables in the white sample.
Household type 4-5, siblings and other relatives, is generally much better situated than any
other household type among the white sample, but not especially so among the black
sample. This is a household type that was noted to be relatively more common among
white owners in Chapter 3. It was suggested that this incidence could be related to the
conservation and transmission of family wealth. This suggestion is reinforced by the
greater wealth of these households, as seen here. In contrast, black households in type 6,
unrelated persons, show greater agricultural resources than many other household types
among the black sample. It should be noted, however, that this finding is largely created
by the one relatively wealth black owning household of type 6, and that the black renting
and sharecropping households of type 6 are not especially wealthier than black renting and
sharecropping households of other types. This is an example of the difficulties in dividing
- the total sample into 30 subsamples.

Household type 7, solitary, almost by definition shows high values in the
categories that are per capita, such as Columns 1, 7, and 9. However, overall, these
households show a lower value of agricultural inputs for the black sample. The white
sample shows much fluctuation, depending on which variable is observed, but overall
shows considerable wealth, especially for one household member. Were variables such as
farm machinery (Column 4), or farm value (Column 5), or traction (Column 2) calculated
per capita, this household type would probably show the greatest wealth in agricultural

inputs of any household type, for the white sample.

Table 5.4 shows the same categories of information broken down by household life
cycle stages rather than household types. These stages are the same ones used in Chapter
3, and are defined by a combination of mother's age (or father’s age if mother not present)

and age of children present in the household. Stage 1 is the youngest stage and Stage 5 is

l S 3
Table 5.3. Agricultural measures by household types, 1880, expressed as means.

 

acres] trac- value value value value valpdl 96 ac ac spot 96 ac
capita tion stock mach farm prod eapita eott Icapita ewpot

(1) (2) (3) (4) (5) (6) (7) (3) (9) (10)

 

 

 

 

 

 

Type Group N

1-2 bl 340 7 1.7 149 36 730 576 109 64 .06 02
1-2 wh 74 16 4.3 423 138 2638 1345 265 59 .12 03
1-2 blow 19 17 3.7 247 48 1479 1036 196 64 .1 06
1-2 blre 161 6 1.7 154 32 635 555 102 6 .08 03
1-2 blsh 160 8 1.5 130 28 697 540 106 68 .03 0.5
1-2 whow 46 20 5.2 521 186 3202 1785 353 6 .15 04
1-2 whre 19 10 3.4 301 60 1255 622 122 54 .09 01
1-2 whsh 10 10 2 212 48 325 605 99 61 .09 01
3 b1 41 7 1.4 127 19 552 484 89 65 .05 02
3 wh 12 16 2.8 362 114 6304 1278 259 49 .14 07
3 blow 2 16 2.5 200 25 1025 576 171 67 .12 01
3 blre 17 6 1.2 132 18 619 504 94 59 .06 04
3 blsh 22 6 1.4 115 21 117 456 78 70 .03 0.7
3 whow 9 20 3.3 422 135 7572 1365 332 53 .15 02
3 whre 3 2 1.3 18 32 600 500 41 31 .12 30
3 whsh 0

4-5 b1 4 5 1.7 106 11 200 609 102 67 0 0
4-5 wh 7 102 8.5 697 375 5407 3442 1309 67 .1 0.1
4-5 blow 0

4-5 blre 2 3 1.5 122 10 200 600 62 56 0 0
4-5 blsh 2 7 2 89 12 0 618 142 76 0 0
4-5 whow 6 17 9.8 800 437 6267 3954 1490 67 .112 0.2
4-5 whre 1 14 1 75 8 250 375 187 71 0 0
4-5 whsh 0

6 b1 6 8 1.6 184 22 1067 632 85 56 .06 0.3
6 wh 4 14 4.5 491 117 1475 1375 141 60 .1 0.3
6 blow 1 3.7 6 700 125 3520 1700 425 67 .5 01
6 blre 3 3 1.7 175 7 250 315 52 57 0 0
6 blsh 2 5 .5 60 7.5 na 200 25 49 0 0
6 whow 1 51 17 1600 450 5000 2500 500 70 .4 0.7
6 whre 1 5 1 125 5 460 250 62 50 0 0
6 whsh 2 0 0 120 6 220 na 0 na 0 na
7 bl 10 23 .9 84 14 227 571 373 72 .20 02
7 wh 5 148 7 461 167 4020 1944 1844 50 .15 .3
7 blow 0

7 blre 6 11 1 90 19 250 324 171 72 .34 03
7 bleh 4 42 .75 74 5 160 901 676 73 0 0
7 whow 3 226 9 7 543 247 6035 2473 2474 59 .17 0.1
7 whre 2 31 3 337 47 650 1150 900 37 .12 33
7 whsh 0

 

 

154

the oldest stage. The sample is again broken down only by race in the first two lines of
each household stage, and subsequently by racial and tenure groupings. Although there
are still problems with low sample N s and empty cells, the breakdown by stage is more
regular than that by household type.

Chapter 3 showed a regularity in household composition, especially regarding the
number of workers, the number of consumers, and the dependency ratio, over the
household life cycle. While the number of consumers most frequently peaked in Stage 3
(with some variation by racial and tenure subsamples), the number of workers usually
peaked in Stage 4. The consumedworker ratio usually peaked, which means was most
”unfavorable,” in Stage 4 also. Most of the variables in Table 5.4 show steady increases
from Stages 1 to 3 (or 4 in some cases), and then decline either in Stage 4 or 5. This is
particularly the case for absolute values like value of livestock (Column 4) or value of farm
(Column 5), which measure absolute wealth, rather than calculated values like percent of
land in cotton (Column 8). Percent of land in cotton shows no clear relationship to
household life cycle, nor does percent of land in sweet potatoes (Column 10). Sweet
potato bushels per capita shows a steady increase over the household stages for the white
sample, but not the black sample. As would be expected, given that households are
smallest at Stages 1 and 5 (see Chapter 3), acres per capita are highest at Stages 1 and 5.

Figures 5.1 and 5.2 illustrate the effects of household life cycle on the value of
standard agricultural inputs like machinery, livestock, land, and the value produced. The
black samples shows a gradual and uninterrupted increase in the value of the farm’s
machinery, and a slight peak and subsequent decline in livestock value. The same pattern
is shown for the white sample, but with some very dramatic increases in Stages 5 for
machinery and Stage 4 for livestock. Farm value increases more rapidly than livestock or

machinery, for both blacks and whites, and then shows more fluctuation between Stages 3

Table 5.4. Agricultural measures by household stages, 1880, expressed as means.

155

 

 

 

 

 

 

 

acres trac- value value value value val pd/ 96 land spot] 96 land
leap tion stock mach farm prod capita cotton capita swpo
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

Stage Group
1 b1 21 10 1 99 16 297 394 136 66 .07 01
1 wh 5 28 3.8 342 84 1200 1098 483 66 .05 0.2
1 blow 0
1 blrc 10 9 1 102 20 142 375 75 60 .1 01
1 blsh 11 11 1 95 9 180 413 147 72 .05 0.4
1 whow 2 21 5 375 85 1500 1283 321 71 12 0.5
1 whre 1 61 3 370 75 600 1300 1300 75 0 0
1 whsh 2 17 3 293 88 381 811 237 56 0 0
2 b1 167 7 1.5 130 25 784 540 94 64 .05 02
2 wh 38 17 4.2 468 145 2519 1387 257 56 .11 04
2 blow 8 29 4.6 196 44 2414 1786 339 57 .11 14
2 blre 70 5 1.4 138 47 639 519 83 60 .07 03
2 blsh 89 6 1.3 116 22 520 441 79 69 .02 0.4
2 whow 23 24 4.8 582 198 931 1845 357 57 .13 06
2 whre 1 7 3.8 335 79 1727 703 118 51 .06 0.7
2 whsh 4 4 1.5 179 29 245 465 71 63 .07 02
3 b1 77 7 2.0 170 42 649 608 94 64 .04 0.6
3 wh 13 12 3.5 366 132 2975 1360 213 62 .14 01
3 blow 4 9 3 219 48 825 625 105 60 .1 0.7
3 blre 38 7 2.1 174 32 581 607 90 64 .04 01
3 blsh 35 7 1.7 156 53 845 607 98 66 .02 0.5
3 whow 12 13 3.7 384 139 3192 1411 221 62 .14 01
3 whre 1 5 2 155 50 380 750 125 62 0 0
3 whsh 0
4 b1 45 8 2.1 167 30 624 726 117 66 .04 03
4 wh 14 12 3.6 4114 151 2754 892 122 62 .16 03
4 blow 4 10 3.3 292 57 867 624 56 80 0 0
4 blre 21 6 2 191 39 554 634 97 58 .06 05
4 blsh 22 10 1.9 123 22 742 806 143 72 .02 0.5
4 whow 8 16 4.7 564 229 636 1290 179 64 .112 02
4 whre 4 6 2.5 231 37 401 233 31 48 .23 05
4 whsh 2 11 1.5 182 10 192 487 73 77 .15 02
5 bl 68 10 1.7 169 32 777 520 141 63 .09 02
5 wh 17 20 4.8 360 1109 5561 1651 444 53 .16 05
5 blow 5 10 2.6 294 43 825 420 129 70 .13 02
5 blre 39 7 1.6 152 56 785 541 139 59 .11 03
5 blsh 23 13 1.6 164 21 200 511 146 66 .06 0.5
5 whow 11 25 6.3 438 139 545 2076 600 56 .19 02
5 whre 5 15 2.4 270 26 633 681 156 50 .1 11
5 whsh 3 16 2.5 227 87 565 792 264 47 .17 0.9

 

 

156

and 5. Value produced shows a peak in Stage 4 for the black sample. The white sample
shows the same pattern, but then a second increase from Stage 4 to Stage 5. Figure 5.3
shows the relationship of traction and life cycle stage for both samples, with the black

sample peaking in Stage 4 and the white sample showing an almost inverse pattern.

Overall, this analysis suggest that some agricultural inputs, such as total livestock
and machinery, generally increased over time and were influenced by the household life
cycle. However, the household life cycle is itself a proxy for time. This means that the
relationship between the household life cycle and these agriarltural measures may largely
be the result of accumulation over time, without regard to household life cycle in the sense
of household composition. Yet many agricultural measures did not show an uninterrupted
increase over time. Those measures which peaked in Stages 3 or 4, with a subsequent
decline, are most likely to show a stronger influence by the fluctuations in household

composition associated with the household life cycle.

Because renters and sharecroppers often provided less of the farm’s machinery and
livestock, and had less of a chance to accumulate these things over time, the relationship
between the household life cycle and the value of these inputs was plotted separately by
tenure. This is shown in Figures 5.4, 5.5, 5.6, 5.7, and 5.8. One observation is that for
all variables, the black sample generally shows a more bell-shaped distribution, suggesting
an increased importance of household composition relative to agricultural wealth in black
households, versus accumulation purely with time. Owners do not necessarily show more
correspondence between the life cycle and these agricultural measures than do renters or
sharecroppers, as was hypothesized above.

157

9.09.. .03. .205 0...: Eocomnoc E 98888 6.5.30.6... ...m 059“.

 

 

 

 

 

 

no... 02? £00.. 02? x03... 02? 30:200.: 02?
0 I ...l o
W m s...
v 08.6 ...... u ..u
"W .W 8n
. ...... m M w
m Wm oov
m 085 I m m oom
. 0006 I m m ooo
con
: oow

158

.0823 .03. .296 8: Eocomaoc 3 $5088 65.50.76,... N..... 0.59...

 

 

 

 

 

 

 

no.5 02? 8.6.. 029 .030 02? Koscuss 33>
v 0 . ll . o

W .8...

HM: .88

m 08.6 U .88

v 3.6 a ooov
m 09.5 m

IL .r 000m
N .85 I

. as... I

159

m as. D
v 085 H
m 088 m
N 035 I

. as... I

Own. .296 0.: Eocmmsoc >9 c0308.. .m.m 0.59m

 

x83

 

 

160

m as. D
V as... E
m as. M
N as. I

. 035 I

 

 

 

 

0.9.0 8.: Eocomsoc .6 53.5008 02m> an. 0.59“.

 

 

 

... .. ... .

 

 

who

 

 

I
‘

 

 

oo—

.om.

oom

omN

161

 

 

m 085 D
v 085 E
m 08.0 W
N 036 I

_ as. I

 

 

0.0.6 0...: Eocmmsoc B 8.8 02; mm 0.59“.

:23 9...; 39.3 cm... 9.2

I . -«.- ,~ ..

 

 

 

 

 

 

 

A

  

.. ooo.

. ooom

. ooom

 

.. ooov

161

 

 

 

m 086 D
v 08% a
m 08.... M
N 085 I

. as... I

 

 

 

 

0.0.6 8.: Eocomsoc S 8.5.. 02m> mm 8ng

39.3 cm... ma...

 

'1 .. .-. -‘rvi‘
i

;. _--. .-

 

 

 

 

 

  

A

.. ooo.

.. OOON

. ooom

 

. ooov

162

0.96 0...: Bocomsoc E 80.68.. 02m> Om 0.59..

:23 0.23 32.3 cm...

 

...n n. ._. ' v'.‘

. oom

 

 

r :75 '7 Jo .. - .
j'jc 7 ......‘_"’-
m cams-7'

 

 
  

 

 

m 003m B .. 000.
v 098m 3

. oom.
m 33m m
m 003m. I .. ooom
. 09.6 I

 

 

 

 

 

 

 

F comm

163

m as... D
v 330 E
n 330 m
N 2030 I

. as... I

 

 

 

0.8.0 0.... 0.050030: .3 .0882. 0:.m>

0...; 30...;

 

__.
xfirxc. - .. . ‘:.

 

 

 

 

 

 

 

...m 8:0:

a... 83

I
.
I
ml
..l
.
i
|
I
..l
..l
_
.l
.a
..l
.
..|
«I

 

 

 

 

.L

l

. oo—

.. CON

.. oov

. oom

 

. 000

.00.»...

I Stage 1
I Stage 2
E Stage 3

m Stage 4

164

D Stage 5

 

 

 

 

 

 

 

 

 
   
 

" .....- .59
_.- .. raft... A. ““

.‘ ' J '7'
,T 34:.“ -.

-

1

”2:. :z'

 

 

 

 

 

 

blre blsh whow whre whsh

blow

Traction by household life cycle and tenure.

Figure 5.8.

165

One final measure which is presented by household life cycle is perhaps the most
basic one. This is ownership itself, shown in Table 5.5 and Figure 5.9. The percentage of
households that owned their farm was very low in the 1880 black sample, but does show a
correspondence with the household life cycle. The same is apparent for the white sample in
1880, although the peak was in Stage 3 instead of Stage 4 and the decline to Stage 4 was
much more dramatic. The differences in the relative incidence of farm ownership by race
were huge in 1880, and still obvious although reduced in 1910. For both blacks and
whites in 1910, the relative incidence of ownership peaked in Stage 3, declined in Stage 4,
and rose back to the Stage 3 level in Stage 5.

These effects are not just those of the age of head (which is sometimes used to
estimate household life cycle). The relationship between age of head and ownership is
shown in Table 5.6. While ownership becomes more frequent the older the head of the
household, the pattern is not the same as that shown by household life cycle. For blacks,
both in 1880 and 1910, there is a gradual and unbroken increase in the percent that own
their farm by age of head. This suggest that while ownership does increase with age,
household composition and its fluctuation over the household life cycle also exerts an
effect, which is to slightly reduce ownership at those points where the number of
household workers is reduced. For whites, the relationship between age of head and
ownership is much more variable. As seen in Table 5.5, the differences between the racial
groups is much reduced in 1910. The steady increase in ownership with age for the black
sample, coupled with a relative decline in the 65+ group of the white sample, resulted in a

level of farm ownership in the 65+ age group that was fairly similar for blacks and whites.

166

Table 5.5. Percentage that own farm, by household life cycle stage.

 

 

 

N 1880 N 1880 1880 N 1910 N 1910 1910
1880 black 1880 white all 1910 black 1910 white all

black %own white %own %own black %own white %own 96own

£839

1 21 0 5 40 8 74 5 10 10 6
2 167 5 38 82 16 214 6 36 36 10
3 77 5 13 92 18 61 18 14 57 25
4 45 9 14 57 20 48 l4 14 50 22
5 68 7 17 58 18 151 18 21 57 23

 

 

Table 5.6 Tenure by age of household head, 1880 and 1910.

 

 

 

1880 1880 1880 1910 1910 1910
N percentage percentage N Wm 960611889
Age Head that rent that own that rent that own
15 to 24 black 84 100 0 70 98 2
white 28 40 60 4 50 50
25 to 34 black 268 96 4 156 94 6
white 59 54 46 20 70 30
35 to 44 black 178 93 7 122 87 13
white 33 19 81 22 45 55
45 to 64 black 467 93 7 152 84 16
white 42 32 65 41 39 61
65+ black 42 91 9 49 75 25
white 5 17 83 3 67 33

 

 

167

.286 8: Eocomsoc 3 :58 E50 “m5 @320an .06. 8:9“.

 

 

 

 

 

92 ..a 99 32; 22 been 82 .3 8o. 32; 82 3.83

l l .l . o

W W W . o.
W “WW W . om
W W . on
.m. M . 2.
m 3.5 D W.W W 1 om
v 086 % W . on
n QDNQW m M I OF
m 085 I W .. on
m . om

_ 093w I

 

noo—

 

The above discussion suggests that household life cycle and household composition
bear some relationship to agricultural wealth and production. To further explore this
relationship, simple regression analysis was conducted, utilizing the basic file of 1880
agricultural census data, organized by households. Because of the high probability of
interdependence between these variables, no attempt was made to conduct multiple
regression. Rather, the focus of the analysis was to compare the relative importance of
each dependent variable in a simple regression equation. Because the preceding analysis
had shown that there were often extreme differences in agricultural wealth and production
by tenure and racial grouping, this analysis was conducted separately for each of the six
racial and tenure groups. In this way, the effects of the demographic or agricultural wealth
variables could be separated from effects of race and tenure. Comparison of the correlation
coefficients is possible in Table 5.7.

Four dependent variables were chosen. The three main dependent variables were
the number of acres cultivated, value produced, and value produced per some unit.
Several different units were used to construct this value produced per unit measure. These
include the number of workers, the number of consumers, and the number of acres,
yielding, respectively, value produced per worker, value produced per consumer, and
value produced per acre. A fourth dependent variable was the bushels of food produced

per consumer.

A series of independent variables were chosen. Each was entered separately into a
simple linear regression equation. The independent variables chosen were those which
had shown variation by racial and tenure groups or household life cycle stage, and which
should, theoretically, have a major affect on agricultural production. They include the

169

demographic variables of the number of consumers, the number of workers, and the
dependency ratio. Agricultural variables include the inputs of machinery, and traction,

measured simply by the absolute number of animals, as traction per worker, and as traction

peracre.

Before discussing the results of this analysis, a caveat should be mentioned. The
regression analysis assumes a linear relationship between the independent and dependent
variables. That is, it assumes increases or decreases of equal proportion in each variable.
To the extent that the relationship between the two variables is not linear, for example, that
an increase in 3 units of the dependent variable does not cause an increase of 3 units but of
only 2 or 1 unit in the dependent variable, the resulting correlation coefficient will be
lowered. In dealing with agricultural production, where there are limits to yields,
relationships may be always be linear. Thus it is suggested that the expectations for the
level of correlation should not be too high and that correlation coefficients of modest

proportions should not be rejected outright.

Initially, it should also be pointed out that there is a strong positive correlation
between the number of workers and the number of consumers in a household. This is
partly because as a person ages, both their working and consuming capacity increases (see
index values used in the Appendix, discussed in Chapter 3). This strong positive
relationship is shown in Column 1 of Table 5.7.

The relationship between the dependent variable number of acres cultivated and the
demographic and agricultural independent variables is generally positive, as would be
expected. The agricultural wealth variables (Columns 4 and 5) show a much stronger
relatiomhip than do the demographic variables (Columns 2 and 3), yielding higher r values
and consistently achieving statistical significance. Except for the black owners, the

170

2.35.. a. :8: 8.3 ..26. 8. 2.. a aegis a £8332 8.8%.: .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.2.. 2.- .2.- .3- .2.. 8. 2. 3. a. .2.. .2. a. a. .2. firfifi
.3. 2.- .8.- .8.- .2.. 8. S. :. 8. .3. .8. .3. «N. .8. 2
.8. 8. 2.- 2. .8. S. 2. :3. 8.- .3. .3. 2. 2. .3. $4.34.
.2. 8.- :x- C.- .2. 2. 2. .2. .2. .3. .R. .8. .8. .2. «1.14.; .
.2. 8.. mm; 8.- .2. .8. om. .3. 2. .S. .3. .8. a. .3. JIIS as
.3. .2. 5.- 2.- .3. 2. S. X. .8. .3. .3. 8. «N.. .3. .65
Idllz :20

ad ad ad ad a a a a g a a a a a

93 :< 59.: =3 :05:
=o=uab 028v own—v cg cocoa-a 3 o 33> an! .6295 33> o 3 o I X
£3

B8 88 U 3 3 vou=v0.qo=_a> v08 snugly—=0 nouod 3

so; coo—evoke o=_a> “IN

 

.8». ...-Es... Basses assess Sea—“E8 .2 2.3

 

171

relationship between the demographic variables and number of acres cultivated is always

positive.

It is important to compare the strengths of these relationships according to racial and
tenure grouping. Notice that the relationship between traction and acres cultivated is
strongest for owners, followed by renters, followed by sharecroppers, for both the black
and white samples. This supports the general understanding that ownership, or at least
access to, traction, was a crucial factor in agricultural production in the postbellum South.
The relationship is probably weaker for renters and sharecroppers because they were more
likely to use animals provided by the landlord, perhaps only intermittently, which might not
be enumerated with that household, and because they probably also had access to lower
numbers of animals per acre. The demographic variable number of workers is more
strongly related to the number of acres cultivated for sharecroppers and renters, although
the correlation coefficients are modest. However, this still suggests that the labor capacity
of the household was an important component in determining farm size. The negative
relationship between number of workers and acres cultivated for black owners may be
bemuse of the age component of land ownership. That is, owners in general were an older
group, or land ownership was more common among older household heads, and more of
these households would have been at a later household stage, when the number of workers
had already peaked and was declining. The same interpretation is suggested for the white
owning sample, but the dampening effect is less strong here (the relationship remains
positive) because white farmers generally did not have to wait as long as black farmers to
own land.

The second dependent variable, value produced, again shows a stronger
relationship between agricultural wealth, here measured by value of machinery, than the
demographic variables. But again all relationships between value produced and the

l 7 2
demographic variables are in the direction expected. It is suggested that value produced per
some unit is a more useful measure. The relationship between value produwd per worker
and the agricultural wealth variable, traction per worker (Column 10) and one demographic
variable, the dependency ratio (Column 1 1), can be compared. Traction per worker shows
a positive effect, which is again strongest for owners. The same holds true when acres
and not workers is used as the unit of measure (Column 14), except that white

sharecroppers show a stronger positive relationship than any other group.

The dependency ratio measures the number of consumers per worker, and is
generally highest during the middle life cycle stages when the household contains more
young children (see Chapter 3). This variable consistently shows a negative relationship
to the value produced per worker (Column 11). This relationship is strongest for
sharecroppers, followed by renters, followed by owners. When the dependent variable is
value produced per acre, not worker, but the independent variable is still the dependency
ratio, a similar ordering is found (Column 13). Owners now show a positive relationship
between value produced per acre and the dependency ratio. Renters and sharecroppers still
show a negative relationship, but it is much weaker. There is no difference in the
relationship between value produced per worker and the dependency ratio (Column 1 l) and
the relationship between value produced per consumer and the dependency ratio (Column
12). This is not surprising given the high correlation between workers and consumers

(Column 1).

The final dependent variable to be discussed is the number of bushels of food
produced per consumer. This variable was discussed above, in relation to agricultural
crop mixes (Table 5.2, Column 2), where it was shown that owners produced between two
and three times the bushels of food per consumer than sharecroppers, with renters showing
an intermediate level of food production. It was also shown that sharecroppers had the

173

lowest percentage of their arltivated land in food crops. One interpretation of these two
patterns is that the higher production of food per consumer for owners was made possible
by their larger acreage. That is, it was hypothesized that all households had to produce a
certain level of cotton, and that production of food crops was then dependent on access to
additional land. Thus the independent variable chosen for the analysis was acres cultivated.
This is shown in Column 6 of Table 5.7. The results are quite mixed. The interpretation
that the number of acres strongly influenced the production of food per consumer is
supported only for the white sharecroppers. While the relationship is generally positive
for other groups, it is weak. A negative relationship is indicated for white owners. This
may be because some of the very largest white owner farms were more commercially
oriented, and produced a relatively low level of food.

The issue of the amount of food the household produced can also be approached
with another type of information from Vinton, which is chattel deeds. The crop lien system
used to provide credit and fumish agricultural producers was discussed in Chapter 2,
generally and at Vinton. The chattel deeds which secured the mortgage on the crop were
also discused in Chapter 2. These deeds do not provide an accounting of each household
at the Vinton store, but they do give some indication of the level of purchases each
household had already made and was expected to make over the year. All chattel deeds to
the Vinton store for the year 1879 were matched to the same households entry in the 1880
agricultural census and household databases. Thirty-two households were matched, but
the analysis was restricted to the 28 households that sharecropped. The 2 owners and 2
renters presented too small a sample for separate analysis, and they were not included with
the sharecroppers because of the differences demonstrated above between the tenure

groups.

174

The information provided by the chattel deeds was the amount of goods which the
Vinton store agreed to advance to that household during the upcoming year. In most cases
this was listed in two steps, the value of goods already provided as of that day, and the
additional value that would be provided over the rest of the season. It is presumed that the
limit of value set by W. E. Trotter was based upon his assessment of each household’s
productive capability, general credit worthiness, and past history of charges at the store. In
several instances the deed indicated that the first entry was not for goods just provided but a
previous balance from the past year, in which case it did indicate a real level of charges.

It was hypothesized that households which normally produced a lower proportion
of food would have a tendency to use buy food from the store, and thus have higher initial
charges and expectations of more charges. The correlation between the total amount to be
purchased that year and the number of bushels of food produced per consumer was in the
expected direction although weak, at -.207. The relationship between store charges and
several dietary staples were also investigated. These variables were the number of swine
on the farm, which yielded a correlation coefficient of -.266, and the total value of all
livestock, which yielded a coefficient of -.286. These results suggest that the level of
purchming at the store was moderately associated with the level at which food was grown
or raised on the farm.

The above examination is tentalizing but difficult to interpret because the meaning of
the value of goods purchased, or allowed to be purchased, in these deeds is unclear.
Much better information on the Vinton store is available from the Watson account ledger of
1911-1912. This ledger and the general nature of its accounts was described in Chapter 2.
To briefly summarize, the ledger gives partial to complete store accounts for about twenty
individuals who were in some way connected with the Henry D. Watson family or Vinton

store. Some are Watson family members, or hired household help, but most are tenant

175

farmers that rented (standing rent, set amount) on the Watson land. They received most of
their goods from the Vinton store, which the Watsons now owned and operated. At the
end of the year they turned their cotton over to the store, and its value was subtracted from
their bill (with interest and other charges, see Chapter 2).

The accounts are itemized and list both the nature of each charge or credit, the
amount, and the date. A total of 555 entries charges were analyzed from these accounts.
Thirty-one percent of these were agriculturally related, and 45 percent were food-related.
The other 24 percent were a mixture of cash, clothes, and miscellaneous expenses like a

doctor’ s bill.

Table 5.8 summarizes entries for 11 black renting households. These eleven
households were chosen because they all had complete accounts and could be matched in
the 1910 population and household files. The first section of Table 5.8 looks at charges
and credits. There was a high level of consistency in the amount each household charged
or the amount it was credited, as expressed in the fairly low coefficients of variability. The
average balance at the end of the year was $- 163.00. Only one household out of the
sample of 11 broke even; they ended the year with a balance of $94. The second section of
the table looks at the most common types of charges. The highest single items were mules
and rent. The average amount spent on food was $49, with meat the single most expensive
item. The average amount spent on agricultural items, was slightly higher at $69. The
agricultural category was defined as any agricultural input, and thus included things like
seed, agricultural tools, and their repair and maintenance. If mules is included with these
other agricultural inputs, the average amount spent on agriculture rises significantly, to
$125. Corn and oats and hay are also major items related to agriculture. It is assumed that
the corn was largely for feed, as meal (undoubtedly corn meal) is listed separately.

Another very significant item was cash advanced by the store, which suggests that these

1 7 6
Table 5.8. Summary of expenses and credits from Watson Ledger (191 l).

 

 

 

 

 

 

ME“ in 8 value ex for coefficient of variation and )
ean inirnurn aximum Coefficient of
Variability“
SUMMARY OF
ENTRIES
all charges 501 209 639 .26
all credits 338 204 462 .21
cotton credit 322 204 462 . 19
balance -163 -323 94 .77
MOST COMMON
CHARGES
corn/oats 27 0 59 .72
hay 7 0 22 1.03
seed 7 0 19 1.18
meat 27 2 42 .43
molasses 3 0 6.5 .84
flour 5 .3 8 .38
meal » l3 0 24 .53
clothes 10 0 52 1.55
cash 59 15 120 .57
all food together 49 5 67 .41
all agricultural 69 30 148 .5
all ag, + mule 125 30 206 .53
interest 64 25 85 .31
rent 112 60 135 .18
PERCENTAGE SPENT
ON
food 10 1.2 20.4 .51
agriculture (+mu1e) 25 08 46 .52
DEMOGRAPHIC
MEASURES
agehead 44 25 70 .3
persons in h 4.4 2 9 .18
dratio 1.03 .87 1.63 .22
credits/worker 134 87 210 .26
credits/consumer 136 68 201 .33
charges/consumer 200 87 340 .39
cotton/worker 128 87 167 .23
food/construct l9 3 37 .52
rent/worker 45 23 60 .31

 

 

 

‘coefficient of variablility calculated as standard deviation/mean.

177

persons were also patronizing other stores, probably in West Point or Aberdeen. Clothes
and other miscellaneous items were less common, and much more variable (see coefficient
of variability). The percentage of total charges that was spent on either food or agriculture
islistedinthethirdsectionofthetable.

Several demographic measures are summarized in the last section of Table 5.8.
The level of variation in the value of credits or charges was low when viewed in relation to
number of consumers or workers. The amount of food charged per consumer is the most
variable (highest coefficient of variability, at .52). No information is available on the
amount of acres these households rented. It is likely that the level of rent reflected the size
of the parcel, although land quality might have played a part. Rent per worker (last row in
table) probably serves as a rough estimate of the acreage available per worker.

Purchases were made throughout the year. Most of the miscellaneous items,
especially clothes, were more common in December, when the account was being settled.
Purchases of agricultural tools were very common in the spring, as the planting cycle was
getting underway. Table 5.9 presents a breakdown of the number of purchases of
agricultural, food-related, and other items by month. Agricultural items here also included
hay, corn, and oats. It can be seen that the number of purchases involving an agricultural
item was fairly consistent throughout the year. Food purchases peaked in August, and
were generally highest in late spring and summer, probably when the cotton crop was
demanding a lot of attention but most gardens were not yet fully producing. Food items
dr0p off entirely in the late fall and winter, when home produce would have still been
abundant.

Because the data on each household’s account could be combined with the person

and household file, it was possible to look at relationships between the household’s store

178
Table 5.9 Schedule of purchases from Vinton store, 1911-1912.

 

 

 

AGRICULTURAL FOOD-RELATED OTHER ‘
Month N 96ofthat N %ofthat N 9601155
month’s month’s month’s
purchases purchases purchases
January 19 61 6 l9 6 19
February 14 35 19 47 7 17
March 16 32 22 44 12 24
April 29 40 34 47 10 14
May 18 23 41 53 18 23
June 20 29 38 54 12 17
July 17 27 32 51 14 22
August 16 29 35 62 5 9
September 8 20 20 51 1 l 28
October 3 23 1 7 9 69
November 3 27 0 0 8 73
December 12 37 0 0 20 63

 

 

 

 

 

Chi-Square computed on counts, significant at .01 level.

account, production of cotton, and household composition. No information on other
crops is available, however, since most of these household did not sell produce to the
store, and since the agricultural census has been destroyed (House of Representatives
1912, 1920). This analysis is similar to that discussed above using the 1880 agricultural
census, with simple regression employed. The sample was 11 households from the ledger
that were matched to the 1910 household frle. Since these households were all black
renters, there is no need to subdivide the sample by tenure. Correlation coeffrcents appear
in Table 5.10.

The analysis of the 1880 agricultural census suggested a weak negative relationship
between acres of land and bushels of food per consumer. Assuming that rent provides an
estimation of acreage, we can look at this relationship in terms of the amount of food
purclmsed. Row 22 in Table 5.10 shows a correlation coefficient of -.46 for these two
variables. The demographic variable of dependency ratio (dratio) does not show much
association with percent spent on food (row 20). The lack of any relationship between

179

Table 5.10. Correlation coefficients, Watson ledger (1911).

 

 

 

Y X R

1 total drarges dratio .39
2 total charges number of consumers .47
3 total charges number of workers .45
4 total charges rent -.85"
5 total charges per consumer dratio -.51
6 total credits rent -.72"
7 total credits percent spent on agriculture .004
8 total credits dratio -.27
9 total credits per worker percent spent on agriculture .55“
10 total credits per worker dratio -.32
11 cotton per worker dratio -.39
12 cottonperwcrker percentspentonagriculture .61"
13 balance owed (as positive number) percent spent on food .56“I
14 balance owed percent spent on agriculture -.18
15 balance owed credits per worker -.34
16 balance owed total credits -. 16
17 balance owed total charges .85‘
18 balance owed dratio .27
19 balance owed rent .57
20 percent spent on food dratio -.03
21 percent spent on food rent -.46
22 percent spent on agriculture dratio -.45
23 percent spent on agriculture rent -.09

 

‘ significant at .05 level
” significant at .10 level

 

rent (as proxy for size of farm) and the percent spent on agriculture (row 23) is surprising.
It may stem for the broadly defined nature of the agricultural category, which mixes
different types of agricultural inputs. The negative relationship between dependency ratio
and percent spent on agriculture (row 22) is interesting, and suggests that households in the
middle stages, when the dependency ratio was highest, may have been less able to direct

resources to agricultural production.

The remainder of the analysis is confined to summary measures from the ledger,
such as charges, credits, and final balance, in an attempt to try to sort out what variables
most strongly influenced the household’s overall success. There is a consistent and

moderate positive relationship between total charges and number of people of different

180

categories in the household, as well as the dependency ratio (rows 1, 2, 3). The strong
negative relationship between total charges and rent (row 4) is very interesting, and
suggests that the larger farms (assuming rent is correlated with size) may have been more

self-sufficient.

Total credits includes cotton credits, which is usually the only credit for most
households. Again there is a strong negative relationship with rent. A more immediately
obvious expectation is that as farm size increased, so would cotton production. However,
this would logically follow from the negative relationship between charges and rent
mentioned above, indicating a lower involvment of some farms with the store, which was
manifest both in lower store charges and lower cotton turned in. There is also a moderate
negative relationship between credits and the dependency ratio (row 8), suggesting that the
productivity of the household is lowest in the second and third stages, when there are
relatively more dependents (see Chapter 3). When total credits is viewed per worker, we
see a similar relationship with dependency ratio (row 10). This relationship holds also
when credits is replawd with cotton production (row 11). Percent spent on agriculture is

positively related to credits per worker (row 9), as is cotton per worker (row 12).

The frnal tally of the household’s year is seen in the balance owed. All but one of
the households owed money at the end of the year. This balance is treated as a positive
number in the analysis. Rent, as proxy for size of farm, is positively correlated with a debt
at the end of the year (row 19), perhaps because of the high negative association between
rent and total charges (row 4). Thus it may be of more interest to concentrate on the nature
ofthe charges. It can be seen in rows 17 and 18 that total charges is more successful than
total credits in explaining the variation in the overall balance, although the relationship of
both variables is in the direction expected. The percent spent on food is positively

associated with a debt at the end of the year (row 13), and percent spent on agriculture is

181

negatively associated (row 14). This also suggests that farms buying more food at the
store were likely to have a higher debt. Although cotton remained the main crop for several
more decades, soils in Clay County were by this time very worn. Also, markups on food
prices were very costly (see Chapter 2). Above it was noted that there was a positive
relationship between percent spent on food and the dependency ratio (row 23). The
relationship between balance owed and dependency ratio is much smaller, but still positive
(row 18).

As mentioned above, only one household broke even at the end of the year. This
was the household headed by Pearce Mealor, composed of Pearce, age 55, and his wife,
and four children. Two of these children were sons age 17 and 19, and two were
daughters age 8 and 1. Interestingly, this was the only household of the 11 to include two
teenage sons. The dependency ratio of this household was an even 1.00, compared to the
mean for the sample of 1.03. In looking closer at the production and charges of this
household, several things stand out. One is that they did not generate the positive balance
at the end of the year by producing relatively more cotton. The Mealor household brought
in $80 worth of cotton per household worker, compared to the sample mean of $128.
Also, they purchased less food. Their value of food purchases per consumer was $15.75,
compared to the sample mean of $ 19. The Mealor household also had very low
miscellaneous charges, and no clothes. Of course this is not to say that these items were
not purchased; but they were not purchased from the Vinton store. Their total charges of
$309 was considerably less than the mean of $50 1 , despite the large household. Their
credits were $403 compared to the mean of $338.

The total credits of Pearce Mealor’s household were unusual because they included
$72.75 worth of labor for the Watson family early in the year. Entries for $15 of credit

were entered on February 1, March 1, April 1, and May 1. On the 25th of May, there was

182

a credit for $12.75 with a note that Pearce stopped work. It is likely that he stopped
working for the Watsons because by May to tend his own crop. However, even without
this $72.50 the Mealor household would have had a positive balance of $21.50. This
household also had one other very unusual entry in the Watson ledger- an entry for $43.80
received on 80 acres of land (see Chapter 2). There also was no rent entry for this
household, a big reason their charges had been so low. If one subtracts the $43.80 land
payment an substitutes $1 12, the mean rent payment, their balance would still be positive,
although much lower ($25.80).

The circumstances surrounding this purchase are not known. Whether this
household was able to buy land because they were able to beat the odds and come out with
a positive balance in previous years, as they had this year, or whether they were able to buy
land because of some special relationship with the Watsons is not known. One other
tenant in the sample also had a credit entry by working for the Watsons. This entry was for

$21.50. However, this household still ended up owing $95 at the end of year settlement.

CHAPTER 6
CONCLUSIONS

The previous chapters presented the findings from the analysis of household
composition, marriage, fertility, and agricultural production. This chapter integrates the
highlights of these findings, especially as they relate to the theoretical perspectives outlined
in Chapter 1. These perspectives include theories about proto-industrialization and semi-

proletarianization, postbellum tenancy, and familylimitation.
Ma '3 e th ousehold Unit

This study showed that marriage was a very common experience for adults in
Vinton. In fact, all black adult women in Vinton had been married at least once, and only a
very small proportion of other samples remained single. There were important differences
between the black and white samples, especially in the higher incidence of serial marriage
and widowhood for blacks. The result was that about one quarter of all black households
included a spouse who had been previously married. The importance of marriage, and the
formation of nuclear family households, was reinforced by the development of a

household-based form of tenant farming.

The analysis of age at marriage showed dramatic differences between owners and
renters. Renters married earlier than owners by 2.35 years for black women, 3.04 years
for white women, 3.98 years for black men, and 2.13 years for white men. Similar

findings were reported by Tolnay (1984). Tolnay’s analysis, which utilized public use

183

184

samples of the 1900 census, differed from the present analysis in that land tenure was not
known for each individual in the sample. Instead, Tolnay calculated singulate mean age of
marriage for different regions, and compared regions with a high level of tenancy with
regions having a low level of tenancy. He found that high tenancy areas showed a
singulate mean age at marriage that was 2.47 years younger for men and 1.45 years
younger for women, in comparison to low tenancy areas. He also found tlmt the ages at
marriage in the rural South at the time were younger by about one year than the urban
South, and younger by about two years than the rural north and west.

Such a differential will likely be interpreted by other scholars as stemming from the
rural nature of the South. I suggest that it may also be influenced by the organization of
Southern agricultural production, in a manner similar to that experienced in the 19th and
19th centuries in areas undergoing proto-industrialization. This analogy is explored more

deeply below.

Age at marriage has long been seen as one of the most important variables in the
demography of the West. The pre-industrial Western demographic regime was
characterized by a close correspondence between marriage and the availability of
agricultural holdings, jobs, or other means of setting up and supporting a separate
household. It resulted in late age at marriage and high levels of celibacy compared to most
other parts of the world (Hajnal 1982). Since the 14th or 15th century, and possibly
earlier, wages or grain prices were the most common variables to which Western marriage
rates and age at marriage responded (Goldstone 1986; Wrigley and Schofreld 1981). Until
the late 18th to 19th centuries, a fairly slow growing economy and an overwhelmingly
agricultural orientation restricted the number of supportable households. This began to
change with prom-industrialization, which occurred in various areas of Europe from the

16th to 19th century, and involved a growth in rural industry, often organized as a putting

185

out system controlled by merchant capital. Proto-industrialization eased restrictions on
household formation. This resulted in an earlier age at marriage, which increased the years
a woman was most at risk of conception, shortened the time between generations, and
stimulated population growth. Much of the population growth was absorbed by the
expanding economy (Archetti 1984; Braun 1978; Kriedte 1981; Levine 1977, 1984;
Medick 1976; Mendels 1972; Tilly 1978, 1984). For example, Palmer (1984) suggests
that the dramatic increases in 19th century European commodity production were
encouraged by the growth of labor as much as by increases in per capita output. The fact
that age at marriage dropped in areas where rural industry was strongest had been noted in
studies since the late 19503 and 19603 (Braun 1960, 1966; Chambers 1957; Klima 1965
cited from Tilly 1984, as well as Levine 1977), but these observations have been hard to

interpret in light of the dominance of demographic transition theory.

The above description is admittedly oversimplified. The growth of rural industry
did not naturally evolve in the countryside, nor, as Mendels points out, was it always
stable. It involved important departures from many long standing traditions like customary
rights of craftsmen, artisan’s guilds, access to common lands, gleaning of agricultural
waste lands, outside service for adolescents, scheduling of traditional holidays, and others
(Blum 1978; Brenner 1976; Kriedte 1983; Mendels 1972). It was encouraged by the
growth of a world economic system and market, as well as rising home demand (Braudel
1981; Schiel 1984; Wallerstein 1974; 1980).

One of the most important aspects of these changes was the articulation of a family
economy, or what several authors have called the Domestic Mode of Production, with rural
industry. This interaction, which will be discussed in more detail below, allowed for rapid
capital accumulation by merchant capitalists, and helped finance the take-off of full
industrialization.

186

Medick (1976) has most fully explored the relationship between the Domestic Mode
of Production (the DMP) and the growth of rural industry. Drawing on Sahlins (1972) and
Chayanov (1966), Medick has argued that the prom-industrial economy was built around
the household, as a self-regulating economic unit that pooled resources and oriented work
toward meeting self-defmed needs. The DMP household was not generally oriented
toward accumulation or profit motives. Medick especially drew on Chayanov’s concept of
self-exploitation, in which the household would intensify labor production when the
dependency ratio or other conditions demanded it, and relax production when needs were
more easily met. It was this history of willingness to ”self-exploit” that made households
more easily manipulated by merchant capitalists, through increases in rents or prices
charged for supplies, decreases in prices for finished goods, demands for work speed-ups,
and so forth. In some cases merchants did not even have to supply enough returns to
support the maintenance and reproduction of the labor force, since rural prom-industrialists
often seemed willing to subsidized merchant capital by side production of a small garden
and a few animals. Compared to previous situations in which a family might unable to
support all children or make use of their labor, and in which many children from poorer
families were sent to work as servants on larger estates, proto-industry increased the
significance of the household as a productive unit and created additional opportunities for

family support.

The importance of the household as a labor unit, its articulation within a large
capitalist socio-economic system, and demographic ramifications, are a useful perspective
not only for proto-industrialization in the West, but also in the contemporary world.
Within the contemporary setting, these processes are often discussed under the rubric of
semi-proletarianization. For example, Wallerstein, one of the leading proponents of this

viewpoint, has described semi-proletarian households as permitting the wage-employment

187

of some of their members at wages below the proportionate cost of reproduction by pooling
wage-income with subsistence agriculture, petty commodity production, and rental and
transfer income (Wallerstein 1983: 18- 19) He, along with Smith and Evans (Smith, Evers,
Wallerstein 1984:8), further describes households as ”systems that are able to provide labor
to capital precisely because they ensure the combination of income from wage labor with
that from non-wage labor so as to form an adequate pool of resources guaranteeing the
replenishment of labor power. Apparently non-capitalist relationships and processes are
responsible for a substantial supply of the world’s labor force and thus are at the very heart

of capitalism.”

The semi-proletarian or proto—industrial interpretive framework has not been applied
to the postbellum South, although Tolnay (1984) has noted the importance of land tenure in
postbellum demographic patterns. However, similarities between Southern tenant farming
and proto-industrial production are striking. For example, in the rural weaving industry of
the 17th and 18th century Europe, it was common for a merchant to supply all of the means
of production, including the loom, and yarn and other supplies, to the working household.
The weaving household basically provided labor and needed to pay loom rent and supply
charges from the returns of their finished product. This is strikingly similar to the case of
Southem tenant farming, where the tenant household often rented all of the means of

production, and was paid out the retums from the crop.

On the macroeconomic level, Southern cotton production produced the raw
materials for Northern and European industrial textile production, and remained a very
important part of the U. S. economy until about 1930, when other fibers increased in
popularity and other countries began to produce more cotton. For example, in 1884,
cotton accounted for 247 million dollars out to a total United States export value of 832

million dollars. Some scholars have suggested the southern economy and interaction with

188

Northern states during this period resembled that of a distinct and dependent colony
(Hacker 1970; Palmer 1984; Woodward 1951). Even the economist Gavin Wright, who
does not fully support the colonial interpretation, has stated that ”the defining feature of this
[the United States’] economy in the post-Civil War era was that the South constituted a
separate regional labor market, outside the scope of national and international labor markets
that were active and effective during the same era” (l986:7). He also noted (1986: 10) that
”southern resource allocation was primarily determined in relationship to the international

economy.”

This relationship can be seen in many statistics. After the Civil War, much
Southern land was under state and federal ownership, through confiscation or because of
owners’ inability to pay taxes. Although some reconstruction programs intended to
distribute such land to ex-slaves, who could not afford to buy it, these programs largely
disolved and the lands were sold, often to Northern industrial interests. Over 5 million
acres were sold between 1377 and 1333. In Mississippi, 339,359 acres were sold to 32
Northern buyers, while only 134,270 acres were sold to 11 Southern buyers. Many
European, especially English, buyers also bought Southern lands during this period. In
1881 in Mississippi, English buyers bought 1,300,000 acres of delta land. Agriculture
became more oriented toward cotton production, with subsequent declines in subsistence
crops and increased reliance on imported food and other basics, and tenancy increased (see

Chapter 2).

Industry also showed the signs of a colonial-like relationship. Many of the more
important railroads, such as the Louisville and Nashville or the Southern, were owned or
controlled by New York or London companies. These companies drew profits out of the
area and protected non-Southem industry by imposing unfavorable freight-rate differentials
upon Southern lines (Woodward 1951:117, 126, 292; Wright 1986.). Industries were low

189

and the Southern cotton was usually exported as a raw material. For example, in 1882,
Southern mills only consumed 942,144 bales out of the 6,949,756 produced (Chew 1884).
Even the cotton mill industries which increased in the South during this period were often
controlled by out-of state corporations and specialized in the manufacture of yarn and rough
cloth. Final processing, in which most value was added and from which higher wages and
other returns were generated, was conducted in Northern or European factories (Hammond
1987; Woodward 1951). In 1910, 62 percent of Southern wage workers worked in the
lower paying extractive industries, compared to 10.7 percent of New England wage
workers, or 14.2 percent of Mid-Atlantic wage workers (Woodward 1951:308-310).

The result was a situation in which accumulation of wealth was directed away from
the Southern states themselves. By 1920, only 42 percent of the non-agricultural wealth of
Mississippi was owned by Mississippi companies (Wright 1986:63). Also, the level of
individual wealth was comparatively low. Per capita wealth in Mississippi in 1880 was
8286, compared to a U.S. average of $870. The lowest per capita wealth of all non-
Southem states was $577 for Kansas, higher than any Southern state (Woodward
1951:111).\\

The present study provides data on accumulation on the local level by two Vinton
merchants, W. E. Trotter, and to a lesser extent, Henry D. Watson. Although agricultural
data was not available for 1910, so that changes in agriculture between 1880 and 1910 are
not available, the study did demonstrate that tenancy increased during this time, and that
tenants were generally less able to devote acreage to foodstuff. They were thus less able to
provide for their own needs, and more reliant on the commission merchant , who in turn
could demand more cotton production. That capital accumulation was quite high was
demonstrated by Trotter’s land holdings and the Dun and Co. financial reports. This

accumulation was based upon the production of cotton by individual households who

190

received a low rate of return for their labor and supplemented these retums with added
contributions from subsistence-oriented farming. However, the local merchant W. E.
Trotter did not display evidence of extreme wealth. The Vinton store did contain a piano,
somewhat of a luxury item (see Chapter 2), and Trotter’s house was undoubtedly one of
the best furnished in the area. However, Trotter‘s house appears to have been a relatively
modest structure (McClurken and Anderson 1981). Chapter 2 demonstrated that his
financial empire was easily destroyed, and eventually taken over by a regional wholesaler.
It is almost certain that a large percentage of the wealth realized at the Vinton store was
clnnneled to the wholesalers and others from whom Trotter obtained goods and credit.

The younger age at marriage for Vinton tenant farmers is similar to the younger age
at marriage for prom-industrialists. Because the present study does not just look at age at
marriage, but also at household composition, it is possible to further explore the household
and its insertion within this form of production. The analysis of household composition
showed the overwhelming dominance of the nuclear family household, both in 1880 and
1910. However, between 1880 and 1910 the nuclear family became even more dominant,
with a resulting decrease in the number of non-nuclear family members within Vinton
households. Historical research has not indicated an obvious reason for this trend.
However, the discussion of agricultural production in the postbellum period presented in
Chapter 2 suggested that this period saw the gradual establishment of household-based
tenant farming. This was discussed in terms of current understandings of Southern tenant
farming, and by changes seen within the project area. For example, there were probably
many persons like Sam Smith discussed above, who did not farm as part of a nuclear
family, in 1880. Certainly there were many more persons living with a family other than
their own compared to the 1910 data set. By 1910, the household-based form of farming
was more common. This may have created additional incentives and pressures for

remaining with a family of orientation or procreation.

191

Thus an interpretation based upon the similarity to the prom-industrial and semi-
proletarian household helps explain a trend which is otherwise difficult to explain. One
alternative explanation would be that black households became more nuclear over time
because of increased assimilation of Euro-Americans kinship. However, there are
problems with this argument. On the one hand, previous historical studies have
documented the prominance of the nuclear family household during and immediately afier
slavery (Gutrnan 1976). Also, the trend of increased nucleation was observed for white as
well as black households in the present study. Several studies of nineteenth century
Southern farmers have documented that white farmers were, like blacks, increasingly
drawn into household-based tenant farming during the postbellum period (Hahn 1983;
Wiener 1978). The present study showed that from 1880 to 1910, tenancy also increased
among white Vinton farmers. It is suggested that the household became increasingly
important as a unit of production in the white papulation as well as the black population,
which encouraged household nucleation.

It is also interesting to note that the category of servants decreased to zero in 1910.
Although it was suggested that this could result from a peculiarity in the census
enumerator‘s use of categories, the trend bears a striking similarity to the decrease in
service among prom-industrial families, who increasingly needed the labor of children and
could support them more easily than before the spread of rural industry. All ethnographic
accounts of the postbellum rural South suggest that children were important sources of
labor in cotton production (Davis et al. 1942; Johnson 1934; Jones 1985; Lewis 1955;
Powdermaker 1939; Rosengarten 1974; Tolnay 1986), as they are in most agricultural
economies (Johnson 1984).

192

While the increasing centrality of the household as the unit of production
encouraged individuals to remain in nuclear family households, the conditions of tenancy
set modest standards, often nothing more than a landlord’s promise of furnishing, for the
formation of a new household. Landlords preferred to rent to families, not single
individuals, which also encouraged marriage and new household formation. The analysis
of individuals’ position within households suggested that tenants did leave the parental
household earlier than owners. This was supported by at least two patterns, in addition to
the earlier age at marriage for tenants mentioned above. One is the higher incidence of
people in their late teens and early twenties residing with parents in owning households.
The second pattern is that more young renters than owners were heads of their own
households. People who left their parents household as young adults did not necessarily
have to form their own household. They could have attached themselves to another
household as a servant, or worked for some other persons or household as a wage laborer.
However, the higher incidence of young household heads among tenants, compared to
owners, suggests that in the postbellum South, they often did form a new household. The
analysis of household composition showed that the stem family household, which has been
associated with the inheritance of property (Berkner 1972) was rare. It was most common
for white owners however, supporting the interpretation that this household form was

encouraged by property transmission.

E I'l'l

Most previous work on interrelationships between proto-industrialization and
demographic processes has focused on age at marriage. Medick, Mendels, and others do
not specifically rule out changes in fertility, although most authors give it little attention. If
anything, they note that the need for labor, and the fact the household was its own labor

unit, did not discourage high fertility. It would be interesting to ask whether abortion,

193

child abandonment, or other signs of unwanted fertility were lower among proto-
industrialists than among similar populations lacking rural industry. Levine (1984) has
also made the suggestion that changes in work routines could have disrupted breastfeeding
patterns, leading to reduced lactational ammenorhea, and increased birth rates.

Despite suggestions that disincentives to high fertility decreased as the household
became a tighter unit of labor, fertility during this period has generally been characterized as
”natural,” and not subject to deliberate manipulation, especially in regard to numbers of
children born. The investigation of fertility undertaken in this study supports this
interpretation, especially for the rural sample. Levels of fertility are relatively high but do
not show the typical pattern of reduction at higher ages (and parities) that has been
associated with the onset of family limitation. This is suggested by the moderate m values
calculated, m being a relative measure of the amount of family limitation in comparison
with a series of mtural fertility populations. Higher m values indicate greater family
limitation.

Results similar to those calculated in the present study have been obtained by
Tolnay et al. (1982), who used the public use samples of the census to look at Southern
white fertility. For example, they found m values for 1886-1889 of .971 urban, .576 rural
non-farm, and .366 rural farm. These values compare to m values in the present study of
.642 for urban owners, .35 for urban renters, .581 for rural owners, and .396 for rural
renters. It should be noted that the West Point sample of the present study is probably less
urban than Tolnay et al.’s urban sample, which would have included more households
from major cities. This may partly explain the higher m values they obtained from their

urban samples.

194

Tolnay (1987) subsequently conducted a similar analysis for another sample, which
included blacks and whites, for 1905-1910. He found m values of .421 black rural non-
farm, .531 white rural non-farm, .338 black rural farm, and .342 white rural farm. This
suggests moderate limitation for the rural non-farm group, and very moderate (if any)
limitation for the rural group. The white samples generally indicate greater limitation than
the black samples. Tolnay notes that as predicted by the historical experiences of Europe,
declines in fertility from 1910 to 1935 were strongest in the older ages. Earlier he had
noted that declines in the 18803 were equally strong in all ages, which is not in keeping
with the usual pattern of family limitation. Tolnay also points out that while m values
suggest rural black fertility was natural, or nearly so, in 1910, fertility had already declined
somewhat from previous levels (see also Chapters 1 and 4), and was on the verge of a
more rapid decline. For example, despite a relatively low m value of .421, which suggests
nearly natural fertility, Tolnay et al. found a low marital fertility rate of 4.760 for the rural
non-farm population. The rural farm marital fertility rate was much higher, 7.425
(compared to the Vinton rate of 6.93, see Chapter 4). This suggests that factors besides

family limitation were reducing fertility, especially in the non- farm population.

The study of marital status presented in Chapter 4, and the analysis of numbers of
children born by different marital histories, suggests that a higher incidence of marital
disruption may have slightly lowered the total fertility rate and the marital fertility rate for
black women. The marital fertility rate is affected because a higher percentage of black
married women probably have more complex marital histories, and more time in between
unions, than white women (see Chapter 4). However, black women were more likely to
be temnts, and it has been shown that tenants married slightly earlier. This could in part
make up for later marital disruption. Also, the study of illegitimacy showed that
childbearing was not as firmly embedded within marriage for the black sample. These

factors combine to suggest that other forces must have also worked to reduce black fertility.

195

Subfecundity and sterility have been noted as important components of black
fertility in the early twentieth century, and may have also been a factor in the late nineteenth
century. Demographers studying Southern black populations are still divided over the
level of subfecundity and sterility, and its relative contribution to fertility decline. The
causes for subfecundity and sterility are also not fully understood, although venereal
disease and general poor health, especially as associated with rickets and increased post-
birth infectious complications, have been the most commonly cited problems (Cutright and
Shorter 1979; Farley 1970; Wright and Pirie 1984). Some of the older writings about the
black fertility decline have focused on the degree to which the black experience followed
the white experience, or the extent to which it presented a unique case (Engerman 1978).
The extent to which family limitation has been underestimated has also been an important

issue (Masnick and McFalls 1976; McFalls and Masnick 1981).

Since Farley (1970), the study to most strongly suggest major differences between
the black and white experience (despite many outward similarities) is Wright and Pirie’s
(1984) research on fertility in relation to the distribution of venereal disease. These authors
concluded that 56 percent of the black fertility decline from the late nineteenth century to
the 19403 was caused by subfecundity and sterility. Tolnay (1989) has suggested that this
percentage is too high, and has recalculated the percentage to 28 percent. Even at 28
percent, this research suggests that subfecundity and sterility were important components
of the black fertility decline. This factor could help explain the high percentage of childless
women for the West Point sample (18.6 percent, see Chapter 4). By 1940, childlessness
among black Southern women age 45-49 was 21.8 percent (Tolnay 1983). Also, it was
shown in Chapter 4 that urban children had a lower survival rate than rural children.
Mortality was not investigated in more detail, so that it is not possible to speak of age-
specific mortality patterns. However, it be might useful in future work to consider the

196

extent to which conditions that encouraged subfecundity or sterility also encouraged higher
infant and early childhood mortality in the urban setting, perhaps through low birth weights
or generally weakened infants.

It is possible that the high percentage of childless women in West Point, and much
lower percentage in Vinton (5.9) (Chapter 4), may be exaggerated by the migration of
childless women from the rural to the urban setting. This is not an unlikely situation
considering that household-based tenant fanning, in which children were an important
component, increasingly dominated the rural area. Not only was farming increasingly
organized by households (Chapter 2), but the tendency for non-nuclear family members to
be added into nuclear family households decreased from 1880 to 1910 (Chapter 3).
Residence in other nuclear family households may have become a less common alternative
for childless black women, with more moving into town where employment opportunities

were good.

To the extent that subfecundity affected black women of all ages but had cumulative
effects that were stronger at later ages, it may have increased the intervals between births at
later ages. This would give the same effect as family limitation, and could account for
some portion of the high m values calculated for black women, especially urban black
women, in this and other studies. For example, the highest m value of the 8 subsamples of
this study, .992, was among black urban owners. Of course, even if we reduce this m
value, say by 28 percent in accordance with Tolnay’s estimate that 28 percent of black
fertility decline was caused by subfecundity and sterility, we have an m value of .714.

This is still high enough to suggest some family limitation. Also, owners showed higher
m values than renters in not just the black urban, but also the white urban and white rural
samples, which suggests that there was some real difference in the level of family limitation

according to tenure. The failure of owners to show a higher m value than renters for the

197

rural black sample is not especially troublesome since both owners and renters in this
sample showed an m value barely over zero, indicating natural fertility (Table 4.8).

We then ask why owners showed a higher level of family limitation than renters.
First, major fertility differences between owners and renters should be restated (see
Chapter 4). These frndings were that rural white women had the highest fertility, followed
by rural black women, urban black women, and urban white women. In seven out of
eight possible comparisons by tenure, renters had higher fertility than owners (Table 4.9).
This finding would be expected if they had a lower level of family limitation. The
differences between owners and renters was generally more dramatic in the rural samples.
This also might be expected when considering the increased importance of tenure in the
rural setting. Whereas in the urban setting, tenure gives information on the ability to
consume one of the most highly prized commodities, housing, and to some extent may
serve as a proxy for overall wealth, it is not as intimately connected with production and
livelihood as in the rural setting.

Historical studies of the relationship between land and fertility in the United States
have usually focused on cost of land, population density, acreage, and other factors within
an owning population. These studies have often assumed the possibility of ownership for
most households and have not focused on tenant farmers (Easterlin et al. 1978; Forster and
Tucker 1972; Leet 1975; Yasuba 1962). This is partly because tenancy was much less
important in other areas and in other times, compared to the postbellum South. Also, many
of these early studies suffered from poor data and methodology (Hareven and Vinovskis
1978 provide a critique). Easterlin et al. (1978), one of the more sophisticated studies,
found little relationship between tenancy and fertility. Their major conclusion was that land

scarcity and high land prices led to the reduction in fertility, both by an increase in age at

l 9 8
marriage and by efforts to restrict fertility within marriage, because families wanted to settle
their children on nearby land.

In the present study, the inability of a large segment of the population to own land
evolved into a system of production and reproduction that did just the opposite, that
encouraged marriage and thus increased fertility. Theories which try to explain a universal
relationship between land and fertility are not likely to succeed, because of varying
meanings and effects of land ownership in different settings. This point is illustrated in the
contemporary world by Johnson’s (1984) summary of varying land and fertility
relationships.

Several historical studies of the United States have been able to analyze fertility by
land tenure. These include Katz and Stem’s (1981) and Stem’s (1983) finding of higher
child-women ratios for farm renters in Erie County, New York in 1855 and 1900. Bash
(1955, 1963) reported a similar finding several decades ago, although except for the work
of Stern and Katz and Stern, is rarely cited. These studies are suggestive and intriguing.
Yet, because they do not have data on age at marriage or detailed fertility measures, they are

hard pressed to explain how, let alone why, tenants had higher fertility.

Tolnay (1984) found the rather confusing result of low fertility in areas of lower
tenancy and in areas of high tenancy, with higher fertility in areas of moderate tenancy.
The fertility variable used by Tolnay was the number of children born per year of marriage,
which was .396 for the high tenancy counties, .473 for moderate-high tenancy counties,
.475 for moderate-low tenancy counties, and .362 for low tenancy counties. These rates
are not too different from those found at Vinton, which were .42 for black owners, and .50
for black renters (Table 4.8). For blacks (and whites) at Vinton, owners showed a lower

number of children born per year of marriage than renters. The analysis of marital status

199

suggested that women in black owning households were slightly more likely to have been
married before, and thus to have had a disruption in marriage. Whether this could be
enough to account for their slightly lower number of children per year of marriage, or
whether renters really had a faster pace of childbearing, remaim to be demonstrated. It is
possible that the low fertility Tolnay observed in high tenancy counties was caused by
conditiom of poor health in some of these counties. (Tolnay 1984).

To return to the issue of the higher level of family limitation for owners, the most
convincing work on this subject has been that of Stern (1979, 1983, also Katz and Stern
1981). The bulk of Stern’s work is based upon an examination of fertility differentials
according to ethnicity and occupation in Erie County, New York from 1855 to 1915. He
found that fertility declined most rapidly for professionals and what he called the ”new
business class” of managers and clerical workers, while it increased for unskilled workers.
Fertility declined slowly for the old business class, and even more slowly for skilled
workers. In his analysis for circa 1900 he found a fertility rate of 3.75 for the new
business class, compared to a rate of 4.53 for the old business class, 5.47 for skilled
workers, and 6.21 for other workers. Although Stern did not calculate m values, he noted
that the fertility reduction among whites was largely in the later years, which is congruent

with family limitation.

One can then generalize from Stern’s findings that because this new business class
was strictly an urban phenomenon, fertility in urban settings should be relatively lower and
show a greater incidence of family limitation, which is the case in the present study.
Although the West Point urban sample is not broken down by occupation, it is suggested
that the owners would have a higher preportion of households in the professional and
mamgerial, or ”new business” class, who would practice more family limitation. It is

hypothesized that the renters would have a higher proportion of skilled and unskilled

200

workers, who should show a lower level of family limitation. The level of family
limitation for urban owners, (.992 black or .642 white) is considerable higher than the m
for urban renters (.679 black or .35 white). However, it should also be noted that the
difference between the urban white and rural white samples is small, and that a moderate
level of family limitation is suggested for rural white owners. It is possible that rural white
owners were increasingly moving away from agricultural production, considering the many
problems associated with postbellum cotton agriculture (see Chapter 2). For example, the
percentage of the rural population that was white declined from 1880 to 1910. In this case,
aspirations for children to move into urban managerial and professional occupations may
have influenced the fertility of rural whites, especially owners, in much the same way that it

did the fertility of the new business class in town.

Stern’s interpretation of the decline in fertility and of family limitation begins with
the earlier observations of Banks (1954) that rising costs of education and declines in real
income squeezed the middle classes into fertility reduction. However, Stern reorients
Banks observations away from changes in the conditions or constraints surrounding
consumption, instead focusing on changes in the organization of production and in the
defrnition of class relations. He interprets the new business class as a new class
formation, dependent upon education and the acquisition of information and managerial
skills in a way that the old business class was dependent on inheritance of capital and other
means of production. The new business class also depended increasingly on conspicuous
consumption to set them apart from other wage workers. Difficulties in meeting these two
goals given large parities led to family limitation. Stern also found that among the business
class, those families with children in school had lower fertility. This does not conclusively
establish his interpretation that a desire to educate children led to a reduction in fertility,
since one could argue that it is those families that had less children that were able to save up

enough for schooling. However, it offers some additional support.

201

'Modo io.

The fact that the household operated under a domestic mode of production, and as a
separate unit of labor, is important to the analogy to proto-industrialization. Chapter 2
demonstrated the prominence of household-based production. The analysis presented in
Chapter 5 showed that household composition and the household life cycle did have import
effects on agricultural wealth, and to a lesser extent on agricultural production. However,
the most important variables in looking at agricultural wealth (or means) and production
were those of race, and especially land tenure. In many cases, such as with the variables
of value produced, value of farm, or acres cultivated, land tenure was more important than
race, so that, for example, the greatest wealth and value produced was seen for white
owners, with the next highest wealth seen for black owners, then white renters followed by
black renters, and so forth. This suggests the strength of class as an explanatory variable.
In fewer cases, such as with value of stock or value of machinery, or number of swine,
differences by race were so great that white renters were better supplied than black owners.
One additional pattern was that white sharecroppers were often sufficiently without means
to rank below black owners and renters. This was true for the variable of traction per acre
or traction per worker, or value farm, or money spent on wages. Whereas some variables
showed little difference between black renters and sharecroppers, white sharecroppers were
usually shown to be much poorer than white renters.

Correlations between the labor power of the household and agricultural production
were generally stronger for the black sample, suggesting their increased reliance on labor,
relative to other agricultural inputs. Crop mixes or farming strategies were not easy to
discern. There was little variation in the percent of land devoted to cotton by racial or

tenure groups. However, owners clearly had more acreage, both improved and

202

unimproved. Unimproved acreage was important for hunting, and for keeping livestock,
and the census data demonstrated that owners produced higher amounts of food crops. For
every tenure group, whites also produced more food than blacks. Analysis of the accounts
of black renting households, as recorded in the Watson ledger, suggested tint households
that were able to focus less on cotton, or at least purchase less food from the store, were
more likely to end the season free of debt. This was supported by the regression analysis
and the case study of the Pearce Mealor household.

The analysis suggests that the lower amount of land available to blacks limited their
ability to increase production. Analysis of the relationship of the dependency ratio to a
number of production variables, both utilizing the 1880 agricultural census data and
utilizing the Watson store ledger, produced mixed results. Usually agricultural wealth
variables, such as traction, value of stock, or value of machinery, were more successful
than the dependency ratio in explaining variation in agricultural production. Although this
finding does not support some of the predictions initially set out for this study, it is perhaps
not surprisingly given that Chayanov’s (1966) work assumed the availability of resources
to intensify production . In the present case, additional acres of farm land were often not
available, and renters and sharecroppers, especially blacks, lacked the credit to obtain land,
or sometimes even basic agricultural supplies. Soil quality was very low by this time, so
that it was difficult to increase yields per acre. Adding fertilizer was the mom important
way to increase yields, but fertilizer was often controlled by the landlord. It was especially
hard for sharecroppers to afford or obtain fertilizer, given their lack of credit. The study
did show that owners were more likely to show a positive relationship between the

consumer/worker ratio and agricultural production.

Chayanov did not propose that household composition was the only, or even
necessarily the most important, variable influencing agricultural production. It was

203

household composition, together with other demands on value created, such as loans or
ruts, that jointly determined a curve of marginal utility, which in turn influenced
production (Durrenberger 1984). Chibnik’s (1987) study of Iowa farm owners in 1880
provided limited support that nineteenth century American farm owners did intensity
production in response to a high dependency ratio. He also demonstrated the value of
household composition in explaining variation in production in a variety of settings
worldwide. Other studies, such as Herring (1984) have also demonstrated the value of
looking at household composition in relation to the agricultural production of tenant
farmers. In this use, tenant farmers were shown to produce more per acre than owners.
While from the perspective of neo-classiul economics, they should have produced less per
acre than owners, because they could not as fully enjoy the benefits of the household’s
labor, Herring showed a correspondence between their increased production and higher

consumer/worker ratios.

One limitation with the present study is that the only variable available to measure
intensifiution of production is the final output, or the value produced. Especially in the
use of agriculture, where yields un vary tremendously according to soil quality and
additives like fertilizer, it would be preferable to have a more direct estimate of labor
intensifiution, or the lack thereof, such as from a time-task analysis. It is quite possible
that labor inputs were incrused according to household needs, but that returns to these

inputs were very small or negligible, due to lack of land, fertilizer, and credit.

The decreased ability of tenant farmers in this setting to effectively intensify
production was suggested by the discussion of the development of this particular form of
tenant farming, and its relationship to racial discrimination. In this and other cases, any
emphasis on ”self-exploitation” that tums attention away from larger structural constraints
may be somewhat misluding. Herring (1984) notes that ”the small farmer with many

204

mouths to feed on little land intensifies production, as Chayanov (along with Kautsky and
Lenin) emplnsized, but with very little choice. The ”self-exploitation” is in fact Mm!
exploitation, related to the structure of distribution of holdings, off-farm employment
opportunities, prices of inputs, output and consumer goods, farm taxes, and so on. The
small farmers who engage in ”self-exploitation” did not create or choose that structure,
despite mystifying voluntaristic construction of both Chayanovian and neoclassiul
perspectives” (Herring 1984:145).

Thus it is crucial that any analysis of household composition and its relationship to
production place this relationship in the broader context of a local economy and in turn its
relationship with macro-economic structures, which the present study has aimed to do.
However, this is not to say that the perspective of the household as a self-regulating unit
aimed at meeting its own needs and taking household composition into account has been
shown to be of little use. The correspondence of household compositional, marriage, and
fertility differentials between households according to their structural positions (renters
compared to owners as well as blacks compared to whites) suggests the importance of class
and racial divisions within postbellum Southern society and suggests that household were
extremely signifiunt units that operated within the constraints of the context in which they

found themselves.

By focusing on these households, their structure and composition, their agricultural
production, and the patterns of marriage and fertility behavior that formed them, and by
looking to the individual perspective as well as that of the household, the present study has
tried to shed light on the processes by which structures, especially class structures, were
created and maintained. In this goal, the study has tried to follow Herbert Gutrnan’s
(1987a) plea for attention to ”essential” questions, or what he called ”the Sartre question”

(Gutrnan 1987b). These questions are ones that focus on how social classes have

205

”interpreted and than dealt with changing patterns of economic, social, and politiul
dependence and inequality” (Gutman 1987b:327). The present study did not assumed that
the participants merely reacted to their surroundings, although they certainly operated
within constraints. In the use of black households, racial discrimination and
accompanying violence munt these patterns or constraints could be very powerful.
However, differences between black and white household composition suggested that the
black households drew upon their own kinship system ,with its greater emphasis on an
enlarged kin network and alterante generational bonds, to cope with the constraints
imposed within the postbellum South. In contrast, signs of an enlarged kin network were
weaker among the white sample, and when present, as in the form of stem family

households, oriented more toward property transmission.

Gutman (1987bz327) further suggests that ”Ameriuns, especially in this century,
have associated the esupe from dependence much too narrowly with possesive
individualism. Addressing the Sartre question and reexamining Ameriun working-class
history reveals a shifting tension inside and outside the workplace between individuals and
collective ways of achieving autonomy.” Gutman’s emphasis is especially appropriate to
the study of postbellum Southern society, for it was in large part ex-slaves desire for
autonomy that influenced the adoption of the household-based system of production that
this study has suggested was so influential in shaping not only the economic but the

demographic structure of the South.

APPENDIX

APPENDIX.
CATEGORIES AND CLASSIFICATIONS

1. Household types
1 husband wife
2 husband and wife plus children
3 singe parents and children
4 adult siblings
5 no conjugality but kinship
6 no kinship
7 solitary (single individual)

11. Categories of extension
1 none
2 single mother
3 single father
4 both parents
5 widowed son
6 widowed daughter
7 widowed son and his children
8 widowed daughter and her children
9 married son and wife
10 married daughter and husband
1 I married son, wife, and children
12 married daughter, husband, anc children
13 grandchildren alone
14 combinations
15 grandfather
l6 grandmother
17 two or more grandparents

111. Categories of Expansion
1 none
2 single brother
3 single sister
4 multiple siblings
5 brother and his children
6 sister and her children
7 brother and sister and children
8 married brother and wife
9 married sister and husband
10 married brother, wife and children
11 married sister, husband, and children
12 niece
l3 nephew
14 multiple nieces and nephews
l5 cousin

206

207

16 cousin and children

17 cousin and spouse

l8 cousin, spouse, and children
19 aunt

20 aunt and her children

21 aunt and uncle

22 aunt, uncle, and children
23 uncle

24 uncle and his children
25 combinations

26 relation uncertain

27 grandneice or nephew

IV. Categories of Augmentation
1 none
2 single boarder
3 married boarder and spouse
4 single boarder and children
5 boarder, spouse, and children
6 single servant
7 servant and spouse
8 servant and children
9 servant, spouse, and children
10 servants, siblings
ll boarders, siblings
12 combinations
l3 friend

2 0 8
V. Household stages

Households are coded by the following stages, based upon
combinations of age of mother and age of children, as follows:

Table A. 1. Household stages.

 

Age of Mother
< 25 25-34 35-44 45+
mishildrenixearsl
no kids 21 22 23 24
all kids < 15 31 32 33 34
at lust one child > 14 41 42 43 44
all kids > 14 51 52 53 54

 

In the analysis, Summary Stages 1 through 5 were defined as
follows, with Stage 1 being the youngest and Stage 5 the oldest.

21 +22
31+32+33
34+43

44+52
23+24+53+54

03th—

A condensed three part system was also developed for use when
sample size was low:

1( oung) 21+22+31+32
2 mid) 33+34+42+43+44
3 (mature) 23+24+52+53+54

V. Relatigenscllrip of individuals to head of household, as designated in census
1
2 spouse
3 son or daughter
4 parent
5 sibling
6 neice or nephew
7 boarder
8 cousin
9 kin, but exact relationship unknown
10 servant
l 1 aunt or uncle
12 grandparent
l3 grandchild
14 grandneice

VI. Marital Status
1. lst marriage

2. 2nd marriage

2 0 9
3rd marriage
4th marriage
single
divorced
widowed
other

VII. Status of children in relation to mother
1. unknown, questionable
2. probable connection
3. good connection
4. good connection, and both partners in first marriage

PS9???

VIII. Reasons for status 1 children (from above)
1. father there, but mother not there
2. more children present than mother has surviving, and serial marriage
3. grandchild of head, and mother not present
4. niece or nephew of head, and mother not present

210

VIII. Measures of Consumers and Workers, for ulculation of dependency ratio.
Following Chibnik (1986)

Table A.2 Consumer/worker equivalents

 

 

 

CONSUMERS
Males Females
AGE (years)
70+ .7 . 7
16-70 1.0 .8
11-15 . 7 . 7
6-10 .4 . 1
0-5 . l . l
WORKERS
Males and Females
71+ .3
61-70 .7
21-60 1.0
16-20 . 7
11-15 . 3

 

LIST OF REFERENCES

LIST OF REFERENCES

Adams, WilliamH., editor 1980 flnvegiy flnntafign; Et hngmhne olggy ninjljen an;
W. Resource Analysts, Inc., Bloomington, Indiana.

Adams, William Hampton, Steven D. Smith, David F. Barton, Timothyu B. Riordan, and
Stephen Poyser 1981 rin 3 Mil: isto ssi
W. National Technical Information Service, Wshington, D. C.

Allen, Ruth Alice 1975 The Labor of Women In In; Egodugign 91 Cotton. New York:
Arno Press. Originally published 1933.

Alston, Lee J. and Robert Higgs 1982 Contractual Mix rn Southern Agriculture since the
Civil War: Facts, Hypotheses, and Tests. MW”: 327-353.

Anderson, Michael 1982 family Structure in Nlnetegnth antuiy Lanmhire. Cambridge
University Press, Cambridge.

Archetti, Eduardo 1984 Rural Families and Demographic Behavior: Some Latin
Ameriun Analogies. (ginnariative Studies in Histgm nnd Sgg'egy 26:251-279.

Arcury, Thomas A. 1984 Household composition and change in a rural community,
1900-1980: testing two models. Ameriun Ethnglogig; 1 1:677-698.

Banks, J. A. 1954 s 't and arenth 0d: tn '1 annin
Victgn'nn Middle Class. Routledge and Kegan Paul, London.

Bash, Wendell 1963 Changing Birth Rates in Developing Ameriu: New York State,
1840-1875. e ilba emorial Fu (1 arte 41:161-182.

Bash, Wendell 1955 Differential Fertility in Madison County, New York, 1865. 11!:
'lba 'al u erl 33:161-186.

Berkner, Lutz 1972 The Stem Family and the Developmental Cycle of the Peasant
Household: An Eighteenth Century Austrian Example. Amen'un Higtnn’cai Review
77:398-418.

Bigham, Darrle E. 1979 The Black Family in Evansville and Vanderburgh County,
Indiana, in 1880. Indinnn Maga_zine Qt flistom 75:117-146.

Billingsley, Andrew 1966 The Treatment of Negro Families' rn American Scholarship. In

W, edited by Andrew Billingsley,p p.p 197 215. Prentice
Hall, Englewood Cliffs, N. J.

Blum, Jerome 1978 :1 h; End of the Old Order in Eng] Eu um; Princeton University

Press, Princeton, N. J.

211

212

Bondurant, John H. and Frank J. Welch 1956 Land Tenure and Agricultural
Development in the South. In LnngLIgnm, edited by K. Parsons, R. Penn, and P. Rays.
University of Wisconsin Press, Madison.

Brenda-Format! 1981-1984 WW Harper
and Row, New York (three volumes).

Braun, Rudolf 1978 Early Industrializatioon and Demographic Change in the Canton of
Zurich In BMrisalMMLChansinaEertilinr edited by Charles Tilly Princeton

University Press, Princeton.

Braun Rudo,lf 1966 The impact of Cottage Industry on an Agricultural Population. In
W edited by David Landes. MacMillan, New York.

Braun, Rudolf 1960 InMfinlisigmngimflQlkslgbgn. Rentsch, Zurich.

Brenner, Robert 1976 Agrarian Class Structure and Economic Development in Pre-
Industrial Europe. W 70:30-75.

Burton. Orville Vernon 1985 MW. Chapel Hill:
University of North Carolina Press.

Caldwell, John C. 1982 Wing Audemic Press, New York.

Chambers, J. D. 1965 Three Essays on the Population and Economy of the Midlands. In
W, edited by D. V. Glass and D. E. C. Eversley. Aldine, Chiugo

Chambers, J. D. 1957 ale re -
thngg. Cambridge University Press, Cambridge.

Chayanov, A. V. 1966 W. Edited by D. Thorner, B.
Kerblay, and R. E. F. Smith. Homewood Press, Chiugo.

Cherlin, Andrew 1978 Remarriage as an Incomplete Institution. Arngflmlqnmnlgf
flogging! 84:634-650.

Chew, Morris R. 1884 W. W. E. Stansburg and Col,

New Orleans.

Chibnik,Mic1nel 1987 The Economic Effects of Household Demography: A Cross-
Cultural Assessment of Chayanov’s Theory. In W;
Innsfgnngfign, edited by Morgan D. Maclachlan, pp. 74-106. University Press of
Ameriu, New York.

Clark. Thomas D. 1944 WWW
Bobbs-Merrill Co., Indianapolis.

Clark, Thomas D. 1946 The Furnishing and Supply System in Southern Agriculture
since 1865. Tammi-ton: 22=pases.

Cleland, Charles E. 1983 Merchants, Tradesmen, and Tenants: The Economics of
Diffusion of Material Culture on a Late Nineteenth Century Site. Wan 23:
35-44.

2 l 3
Cleland, Charles E. and Kim A. McBride, eds. 1983 Oral Historical, Mmgim, and

a r inlOuz: luv“ “you. N-rton (1 ° 01 uiissiu” a- I 1'!” ' our 0
0 th b' eHi to' ' ' . Michigan State University

Musuem, East Lansing, Michigan.

Coale, Ansley J. 1973 W. International Union for
the Scientific Study of Population, Liege.

Coale, Ansley 1.1986 The Decline of Fertility' rn Europe since the Eighteenth Century as
a Chapter in Demographic History In W edited by Ansley
J. Coale adn Susan Cotts Watkins, pp. 1-30. Princeton University Press, Princeton.

Coale. Ansley 1- and Paul Demeny 1966 WWW:
W Princeton University Press, Princeton.

Coale, Ansley J. and N orfleet W. Rives 1973 A Statistical Reconstruction of the Black
Population of the United States, 1880-1970: Estimates of the True Numbers by Age and
Sex, Birth Rates, and Total Fertility. BMW 39:3-36.

Coale, Ansley]. and T. J. Trusell 1974 Model Fertility Schedule: L Variations in the Age
Structure of Childburing' in Human Populations. Wig; 40: 185-258.

Coale, Ansley J. and T. I. Trusell 1978 Techniul Note: Finding the Two Parameters that
Specify a Model Schedule of Marital Fertility. W 44:203-213.

Coale. Ansley J and Melvin Zelnik 1963 W
W. Princeton University Press, Princeton.

Cooper, Owen 1933 MW. Mississippi State Planning

Commission Release N o. 7, Jackson, Mississippi.

Cowgill, George 1975 On the Causes and Consequences of Ancient and Modern
Population Growth WW 77: 505-525

Cox, La Wanda 1944 Tenancy in the U. S., 1865- 1900. A Consideration of teh Validty
of the Agricultural Ladder Hypothesis. Agflgfliuflflintgg 18:97-105.

Elder, Glen H. J r 1977 Family History and The Life Course. MW
2:279-304.

Correspondent 1865 W233“.

Cutright, P. and E. Shorter 1979 The effects of health on the completed fertility of
nonwhite and white U. S. women born between 1867 and 1935.1;an
13:191-217.

Daniel, Pete 1984 The Crossroads of Change: Cotton, Tobacco, and Rice Cultures in the
Twentieth-Century South. WM L:29-456.

Darroch, A. Gordon and Michael Omstein 1984 Family and Household in Nineteenth
Century Canada: Regional Patterns. WW 9: 158-177.

Davis, Allison, Burleigh Gardner and Mary A. Gardner 1942 Dggpjmnh. Chiugo
University Press, Chiugo.

214

DeCanio, Stephen 1974 Agriglture m mg Postbeiluni South; 11.19 liming 9!
W. Harvard University Press, Cambridge, Massachusetts.

Demeny, Paul 1968 Early Fertility Decline in Austria-Hungary: A Lesson in
Demographic Transition. Mn: 97:502-522.

Durrenberger, E. Paul 1984 Q: . :
Audemic Press, New York.

Easterlin, R. A., G. Alter, and G. A. Condran 1978 Farms and farm families in old and
new areas the Northern states in 1860 In Eamflxandfionsrlaflonjnuinmmm
Alum edited by T. Harevin and M. Vinvoskis, pp. 22- 85. Princeton University Press,
Princeton.

Eblen, Jack Ericson 1974 New Estimates of the Vital Rates of the United States Black
Population During the Nineteenth Century. ngngmny 11:301-319.

 

Edward, Thomas J. 1913 The Tenant System and Some Changes Since Emancipation.
de olitiul d 49:38-46.

Engerman, Stanley L. 1978 Changes' in Black Fertility, 1880-1940. In Enmflmd
tu edited by T. K. Hareven and M. A. Vinovskis,
pp. 126-153. Princeton University Press, Princeton.

Ferleger, Louis 1984 Self-Sufficiency and Rural Life on Southern Farms. Agcigilinm
11133911 58:314-329.

Farley, Reynolds 1970 $2132th 91 thg Black Ennulation: A Study 91 dgmgginnhig immis-
Chiugo: Markham Publishing Co.

Fite, Gilbert 1986 The Agricultural Trap in the South. 13mm 60:38-50.

Fite, Gilbert 1984 £52an Eield§ N9 More. The University Press of Kentucky,
Lexington, Kentucky.

Flynn Charles L.- Jr 1983 Waist:
W318- Louisiana State University Press, Baton Rouge.

Forster. C. and G. S. L. Tucker 197 2 Warm
Yale University Press, New Haven.

Fortes, Meyer 1971 The Developmental Cycle in Domestic Groups. In W
Raging, edited by Jack Goody. Penguin Books, Baltimore.

Foster, Herbert J. 1983 Afriun Patterns rn the Afro-Ameriun Family. mm
Snidieg 14(2): 201- 232.

Frazier, E. Franklin 1939 jlfhg Negro Enmily in the United Statu. University of Chiugo
Press, Chiugo.

Furstenberg, Frank F., J r Theodore Hersberg, and John Modell 1975 The Origins of
the Female-Huded Black Family: The Impact of the Urban Experience. ananQf

InterdisnnlinamBistamtit-le 233

215

Goldenweiser, E. A. and Leon E. Truesdell 1924 W Census
Monogra phy IV. Government Printing Office, Washington.

Goode. William J 1963 W University of Chicago
Press, Chicago.

Graybill, Wilson H. and L. J. Cho 1965 Methodology for the Measurement of Current
Fertility from Population Data on Young Children. My 2:50-73.

Graybill, Wilson H., Clyde V. Kiser, and Pascal K. Whelpton 1958 Ihgfignjfitygf
Amgflgnflqman. John Wiley and Sons, New York.

Gutman. Herbert 1976 W- New
York: Vintage Books.

Gutman, Herbert 1987a The Black Family rn Slavery and Freedom: A Revised
Perspective. In ' . ‘ _ -- : edited by Ira
Berlin, Pp. 357- 379. Pantheon Books, New York.

Gutman, Herbert 1987b Labor History and athe ”Sartre Question. In W
_ , edited by Ira Berlin, Pp. 326-328.

 

 

PantheonBooks New York. 1‘
Hacker.louisM. 1970 nmmamflrmmmmm

John Wiley and Sons, New York.

Hahn, Steven 1983 I." ; mu ~ .
MW Oxford University Press New

   

Hajnal, John 1982 Taro kinds of pre-industrial household formation systems. Wm
W 8:449-494.

Hajnal, John 1965 European Marriage Patterns in Perspective. In W
D. V. Glass and D. E. C. Eversely, eds, London.

Hajnal, John 1953 Age at Marriage and Proportions Marrying. mm 7:1 1 1-
136.

Hamme1,E A. and Peter Laslett 1974 Comparing Household Structure Over Time and
Between Cultures Wm 16: 73- 109

Hammond, M. B. 1897 MM. Macmillan Company, New York.

Handwerkcr, W. Penn 1983 The First Demographic Transition: An Analysis of
Subsistence Choices and Reproductive Consequences. MW 85: 5- 27.

Handwerkcr, W. Penn 1977 Family, Fertility, and Economics. W
18:259-266.

Hareven.TamaraK editor 1978 W
Wig. Academic Press, New York.

Hareven, Tamara K. and Maris A. Vinovskis 1978 Patterns of Childe in Late
Nineteenth-Century America: The Determinants of Marital Fertility in Five Massachusetts

216

Towns In 1880 In Wmm edited byT K
Hareven and M A. Vinovskis, pp. 85-125. Princeton University Press, Princeton, New
Jersey.

Harris, William 1976 Work and the Family in Black Atlanta, 1880. W
Hm 9:319-330.

Henry, Louis 1961 Some Data on Natural Fertility. W 8:81-91.

Herring, Ronald J. 1984 Chayanovian versus Neoclassical Perspectives on Land Tenure
and Productivity Interactions In Wham.
edited by E. Paul Durrenberger, pp. 133- 149. Academic Press, New York.

Higgs, Robert 1984 The Boll Weevil, the Cotton Economy, and Black Migration, 1910-
1930. MW 50:335-350.

Higgs,Robert 1977
1865:1214. Cambridge University Press, Cambridge.

 

House of Representatives 1912 List of Useless Papers in the Bureau of the Census.
House of Representatives Document No. 460, 62nd Congress, 2nd Session.

House of Representatives 1920 Disposition of Useless Papers in the Department of
Commerce. House of Representatives Document No. 750, 66nd Congress, 2nd Session.

Johnson, Charles S. 1934 W911. Chicago: University of Chicago
Press.

Johnson, Nan E.1984 Rural Development and the Value of Children: Implications for
Human Fertility In W edited by Wayne A. Schutjer
and C. Shannon Stokes. MacMillan, New York.

Jones, Jaccqueline 1985 Lnbnr nf Igve, Labor of Sorrow: Blank annen, W911; and 111:
W. New York: Basic Books.

Katz, Michael B, and Mark J. Stern 1980 History and the Limits of Population Policy.
W 10: 225-245

Katz, Michael B. and Mark]. Stern 1981 Fertility, Class, and Industrial Capitalism:
Erie County, New York, 1855- 1915. W33: 63- 92.

Katz, Michael B., Michael J. Doucet, and MarkJ. Stern 1982 ' '
WM. Harvard University Press, Cambridge, Massachusetts.

Kern, John R., Steven F. Miller, Ira Berlin, and Joseph P. Reidy 1982 Shngnlggs
inn; .istoric 't ° PhaseII ' 9'ca 1V‘t'--ti0 Orr-11'.“ :. ~ Multim. r
Distn'gt, Qabama and M ississinni. Commonwealth Associates, Jackson, Michigan.

Kertzer, David 1984 a u ' ' . ‘
W. Rutgers University Press, New Brunswick, New Jersey.

Kertzer, Davrd I 1985 Future Directions rn Historical Household Studies. anmnlnf
Eamilfliistam 10:98-107.

 
 

2 1 7
Klima, Amost 1965 The Domestic Industry and the Putting- Out System in the Period of
Intuition from Fedualism to Capitalism In W
W pp 477-482 Mouton. Paris

Kirby, Jack Temple 519583 The Southern Exodus: A Primer for Historians. InnrmLQf
Sauthmflistm 49: 8 ~600-

Kolchin, Peter 1972 ~ . 7 ;_ .2- . -
W. Westport, Conn.: Greenwood Press.

Knodel, John 1977 Fertility Limitation and the Fertility Transistion: Evidence from the
Age Patterns of Fertility in Europe and Asia. W 3:481-521.

Knodel, John 1979 From Natural Fertility to Family Limitation: the Onset of Fertility
Transition in a Sample of German Villages. W 16:493-521.

Kriedte, Peter 1983 ' 4;; g .
W. Cambridge University Press, Cambridge.

Kriedte, P., H. Medick, and J. Schlumbolm 1981 mm
mm. Cambridge University Press, Cambridge.

Lantx, H. and L. Hendrix 1978 Black fertility and the black family’ tn the nineteenth
century: A re-examination of the past. W 3: 251- 261.

Laslett, Peter, and Richard Wall, editors 1972 WWW.
Cambridge University Press, Cambridge, England.

Leet, Don R. 1975 Human Fertility and Agricultural Opportunities rn Ohio Counties:
From Frontier to Maturity. 1810-1860 In W

,edited by David C. Klingman and Richard K. Vedder. Ohio
University, Athens, Ohio.

Lesthaeghe, Ron 1983 A century of demographic and cultural change m Western Europe:

An exploration of underlying dimensions. WW 9: 41 1-
435.

Leridon, Henry 1977 Hnman Eertility. University of Chicago Press, Chicago.
Levine, David 1984 Production, Reporduction, and the Proletarian Family in England,

1500-1851. In 1 ' i tioo an amil isto pp. 87-127. Academic Press,
New York.

Levine. David 1977 WWW Academic
Press, New York.

Levine, Robert A. and Susan C. M. Scrimshaw 1983 Effects of Culture on Fertility:
Anthropological Contributions In W.
edited by R. Bulatao and R. Lee, pp. 666-695. Academic Press, New York.

Lewis, Carolyn Baker 1984 The World Around Hampton: Post-bellum Life on South
Carolina Plantation. Agflgnlnnnlflsm 58:456-476.

Lewis, Hylan 1955 Elm. University of North Carolina, Chapel Hill.

 

 

218

Litwaek.LeonF. 1979 WWW Vintage
Books, New York.

Mandle, Jay 1978 W. Duke University Press, Durham.

Manfra, Jo Ann and Robert R. Dykstra 1985 Serial Marriage and the Origins of the Black
Stepfamily: The Rowanty Evidence. WM 72:18-44.

Martin, E. P. and J. M Martin 1978 W1, Chicago: University of
Chicago Press.

Masnick, G. S. and J. A. McFalls 1976 A new perspective on the twentieth-century
American fertility swing. W812!!! 1: 217- 244.

McBride, Kim A. 1983 The Study rn Historical Perspective: Changes and Continuities at
Barton and Vinton, Mississippi. In ral Historical ocumenta and

i sissi ., edited by Charles E. Cleland Iand Kim
A. McBride, pp. 74- 134. Michigan State University Museum, East Lansing, Michigan.

McClurken, James M. and Peggy Uland Anderson, compilers 1981 Mm
' w T ' bi 'sto ' 'tes '

Michigan State University Museum, East Lansing, Michigan:

McFalls, Joseph A. J r., and George S. Masnick 1981 Birth Control and the Fertility of
the U. S. Black Population 1880-1980. WW 6:89-106.

McKendrick, Neil 1974 Home Demand and Economic Growth: A New View of the Role
of Women and Children in the Industrial Revolution. In

W, edited by Neil McKendrick. Europa, London.

McNicoll, Geoffrey 1978 Population and Development: Outline for a Structuralist
Approach. WWW 14(4):78-90.

Medick, Hans 1976 The prom-industrial family economy: structure and function in the
transition from peasant agriculture to industrial capitalism. mm 1:291-315.

 

Meeker, E. 1977 Freedom, economic opportunity, and fertility: Black Amerieans, 1860-
1910. Eggnnmjglnmflm 15:397-412.

Mendels, Franklin 1972 Proto-industrialimtion: The First Phase of the Industrialization
Process MW 32: 241 -261

Mintzo, SidneyW. and Richard Price 1976 h '
, , -. -_ ‘ z- 'v. Institute for the Study of Human Issues,

  

Nag, Moni 1980 How Modernization Can Also Increase Fertility. W
21:571-587.

Nardi, Bonnie Anna 1981 Modes of Explanation in Anthropological Population Theory:
Ecological Determinism Versus Self-Regulation in Studies of Population Growth in Third

World Countries. W983! 83:28-56.

219

Nixon, Herman Clarence 1938 W University of North
Carolina, Chapel Hill.

Notestein, Frank 1953 Economic Problems of Population. In W
W S t e Ei te a Conf re
Lfignnnnljns, pp. 13-31 and 61 -.65 Oxford University Press, London.

Okun. Bernard 1958 W The Johns
Hopkins University Press, Baltimore.

 

Palmer, Bryan D. 1984 Social Formation and Class Formation rn North American, 1800-

1900. In Wm, edited by David Levine, pp. 229- 309. Academic
Press, New York.

Parker, William N. The South in the National Economy, 1865-1970. Wig
1631112146: 1019-1048.

Pearson, Elizabeth, editor 1969 W. Arno Press, New
York. Originally published 1906.

Pleck. Elizabeth 1979 WWW- New York:
Academic Press.

Polgar, Stephen 1972 Population History and Population Policies from an
Anthropological Perspective. W 12:203-21 1.

Powdermaker. Hortense 1939 WWW.
Viking Press, New York.

Rainwater. Lee and William Yancy 1967 W
m. Cambridge: Harvard University Press.

Ransom, Roger and Richard Sutch 1977 MW. Cambridge University
Press, New York.

Raper, Arthur F. 1968
York: Atheneum.

 

Raulet, Harry M. 1970 Family Flaming and Population Control in Developing Countries.
Dunggmzlm 7:21 1-233.

Retherford, Robert D. and Lee Jay Cho 1978 Age-Parity-Specifrc Birth Rates and Birth
Probabilities from Census of Survey Data on Own Children. W 32: 567-
58 1.

Rindfuss, Ronald R. 1981 Fertility. In W edited by D. Postan
and R. Weller, pp. 23-54. University of Texas Press, Austin.

Ripley, C. Peter 1975 The Black Family in Transition: Louisiana, 1860-1865. Ihg
WWW XLI:367-380-

Rosengarten. Theodore 1974 WWW Avon Books.
New York.

 

220 .
Sahlins, Marshall 1972 W- Aldine, Chicago.

SandersJ T 1922 WWW USDA
Department Bulletin 1068, Washington, D. C.

Schiel, Tilman 1984 Development and Underdevelopment of Household-Based
Production in Europe. In HgMnLdsnnd—thgflndnm edited by Joan Smith,
Immanuel Wallerstein, and Hans- Dieter Evers, pp. 101- 130. Sage Publications, Beverly
Hills, California.

Sbifflett, Crandell 1982 31.1-15.1 '
W. Knoxville: University of Tennessee Press.

Sbifflett, Crandell 1975 The Household Composition of Rural Black Families, Louisa
County. Virginia- WW 6: 235- 260

Sbimkin, Demitri B., Eidth M. Sbimkin, and Dennis A. Frate 1978 W11!
mm The Hague, Paris: Mouton.

Shlomowitz, Ralph 1979 The Origin of Southern Sharecropping. Agfigfltnralflistgm
53:557-575.

Shlomowitz, Ralph 1984 ”Bound” or ”Free”? Black labor in cotton and sugarcoane
faring, 1865-1880. WM L:569-596.

Smallwood, James 1977 Emancipation and the Black Family: A Case Study. Sgginl
W 57:849-857-

Smith, Joan, I. Wallerstein, and Hans-Dieter Evers 1984 Introduction. In Hnngghnlngjn
edited by Joan Smith, I. Wallerstein, and Hans-Dieter Evers, pp. 7-
13. Sage Publications, Beverly Hills, California.

Smith, Richard 1981 Fertility, Economy, and Household Formation in England Over
Three Centuries Wm 7: 595-622

Smith, Steven D., David F. Barton, and Timothy B. Riordan 1982 BMW
- ' d 1- - .National Technical

 

 

Information Service, Washington, D. C.

Stack. Carol B. 1974 W New
York: Harper and Row.

State of Mrssrssrppr 1880 Chapter 50, An Act rn Relation to landlord and Tenant. In
W Jackson. Mississippi

Stern, Mark J. 1983 Differential Fertility rn Rural Erie County, New York, 1855. 151nm]
Wm 16: 49- 63

Stem, Mark J. 1979 The Demography of Capitalism: Industry, Class, and Fertility in
Erie County, New York, 1855-1915. Ph.D. dissertation, York University, Toronto.

Stolnitz, George 1964 The Demographic Transition: From High to Low Birth Rates and

Death Rates. In W, edited by Ronald Freedman, pp. 30-46.
Doubleday and Co, Garden City, New Jersey.

221

Strickland, John Scott 1985 Traditional Culture and Moral Economy. In Ihg
‘ -- '111, nun-r editedbyStevenHabnandJohatban
Prude, pp. 141-178. University of North Carolina Press, Chapel Hill.

 

Sudarkasa, N iara 1980 African and Afro-American family structure: a comparison.
mm (November/December):37-60.

Taylor, Rosser H. 1943 Post-bellum Southern Rental Contracts. Wm
17: 12 1 - 128.

Teitelbaum, Michael S. 1975 Relevance of Demographic Transition Theory for
Developing Countries. Sgigngg 188:420-425.

Tilly, Charles 1984 Demographic Origins of the European Proletariat. In

lemrianizstimsnfliamilaflistam edited by David Levine Academic Press. New
York.

Tilly, Charles 1978 The Historical Study of Vital Processes. mm
W, edited by Charles Tilly, pp. 3-55. Princeton University Press,
Princeton.

Tilly, Louise and Joan Scott 1978 MW. Holt, Rinebart, and
Winston, New York.

Tolnay, Stewart 1989 A new look at the effect of venereal disease on black fertility: the
deep South in 1940. Qgrnngmnhy 26:679-690.

Tolnay, Stewart 1987 The decline of black marital fertility in the rural South: 1910-1940.
MW 52: 21 i -2 i7

Tolmy, Stewart 1986 Family economy and the black American fertility transition.
Wm 1 1: 268- 283

Tolnay, Stewart 1985 Black American fertility transition, 1880-1940. W4
mm 70:22-7.

Tolnay, Stewart 1984 Black Family Formation and Tenancy rn the Farm Soutb,1900.
MW 90 105- 325

Tolnay, Stewart 1983 Fertility of Southern Black Farmers in 1900. Wily
Him 8:314-332.

Tolnay, Stewart 1981 Trends in Total and Marital Fertility for Black Americans, 1886-
1899. My 18:443-463.

Tolnay, Stewart, Stephen N. Graham, and Avery M. Guest 1982 Own-child estimates of
U. S. white fertility, 1886-99. W 15(3): 127- 138.

Unites States Census Bureau 1960 W U S
Census Bureau and Social Science Research Council, Washington, D. C.

United States Census Bureau 1933 MW Government
Printing Office, Washington.

 

222

United States Census Bureau 1918 WWW
Government Printing Office, Washington.

United State Census Bureau 1916 WWW. Government
Printing Office, Washington.

United States Census Bureau 1913 W
W. Government Printing Office, Washington.

United States Census Bureau 1884 WW9:
by Eugene W. Hilgard, as part of the Tenth Census of the United States. Government
Printing Office, Washington, D. C.

United States Industrial Commission 1901
W. Government Printing Office, Washington, D. C.

van de Walle, Etienne and John Knodel 1980 Europe's Fertility Transition: New
Evidence and Lessons for Today' 5 Developing World. WM 34: 3- 43.

Verdon, Michel 1980 Shaking off the Domestic Yoke, or the Sociological Significance of
Residence. William 22109 133

Wall, Richard 1983 W. Cambridge University Press,
Cambridge, England.

Wallerstein, Immanuel 1983 W. Sage Publishing Co.,
Beverly, Hills, CA.

Wallerstein. Immanuel 1980 nmmmamm-Ammmmmm
Causalmanansfthafiummnflorldfimamyw Academic Press New York.

Wallerstein, Immanuel 1974

Wm
WM Academic Press. New
York.

Watson Collection ad. Account lodgers. Special Collections, Mitchell Memorial
Library, Mississippi State University, Starkville, Mississippi.

Way, Winston W. and Kim A. McBride 1983 The Study in Historical Perspective: A
Narrative of Mid-Nineteenth Century Barton and Vinton, Mississippi. In W
L0. "1.1: 10.11 sleet 1 ‘ 1011 urn 110 01 u ' ”edited
by Charles E. Cleland and Kim A. McBride, pp. 7- 73. Michigan State University
Museum, East Lansing, Michigan.

Weiman, David F. 1985 The Economic Emancipation of the Non-Slaveholding Class:

Upcountry farmers in teh Georgia Cotton Economy. MW 45: 7 1-
93.

Welch, Frank 1943 MW. Bulletin 385, Agricultural Experiment
Station, Mississippi State University, Starkville, Mississippi.

Wesson, Kenneth R. 1980 The Southern Country Store Revisited: A Test Case.
Alabamaflmnflfluanefly 42.157 166

223

Wharton, Vernon Lane 1965 W. Harper and Row,
New York. Originally published 1947.

Wiener, Jonathan M. 1978 W11. Louisiana SState

University Press, Baton Rouge.

Wilk, Robert and Robert NcC. Netting 1984 Households: Changing Forms and
Functions In WWW.
edited by Robert McC. Netting, Richard Wilk, and Eric J. Amould. University of
California Press, Berkeley.

 

Wolf, Eric 1966 Egaaams. Prentice-Hall, New York.

Woodman, Harold D. 1985 The Reconstruction of the Cotton Plantation in the New
South. In W edited by Thavolia Glymph.
pp.95-119. Texas A and M Press, College Station.

Woodman, Harold D. 1984 How New Was the New South? W
58:529-545.

Woodman, Harold D.1968 KinWmaEinanrinmaMatketing
W. The University Press of Kentucky, Lexington.

Woodward, C. Vann 1951 W. Louisiana State

University, Baton Rouge.

Woofter. Thomasl. Jr 1969 W
W. New York Originally published 1920.

Wright, P. and P. Pirie 1984 A False Fertility Transition: The Case of American Blacks.
W No. 90. East-West Center, Hawaii.

Wright, Gavin 1986 S2151 Snn1h, New $911111. Basic Books, Inc., New York.

Wright, Gavin 1979 Freedom and the Southern Economy. Wm
1115111131 16: 103- 110.

Wright, Gavin, and Howard Kunreuther 1975 Cotton, Corn and Risk in the Nineteenth
Century loumalsLEsanamiaHistm 35: 526-

Wrigley, E. A. 1966 Family Limitation in Pre-Industiral England. Eggnnmigflimm
Rm 19: 82-109.

Wrigley.E A.. andR S Schofield 1981 W
W Amould. London

Yasuba. Y. 1962 WWW
Johns Hopkins University Press, Baltimore, Maryland.

 

 

I'.flllllulll "lit-u ... t