' Ni |-“J _',¢ 6 . 3. -' ,..-‘ .
‘-":?3‘:3g;:3’_}.§§§’?‘;c}'s 'y _. :- .- a. oA 'J . f;

. . _ n. . “,wfirn asuea-N'vx‘uwv‘fv‘flv'movdflnqwm

 

POPULATEON CHANGE IN CENTRAL SOUTH-MAMPRUSI,
NORTHERN GHANA ~

Thesis for the Degree of M. A.
MICHIGAN STATE UNIVERSITY
JOHN SEBIYAM NABILA

‘ ' 1970

 

 

 

 

\IUIHWH“\||H\|l\\“|\“\|\|\\HHHIHHIJINMHW ,1“wa g;—

3 1293 10437

 

ABSTRACT

POPULATION CHANGE IN CENTRAL SOUTH-
MAMPRUSI, NORTHERN GHANA

by John S. Nabila

Population change is a dynamic equilibrium between forces of
increment and forces of decrement. Thus, change in the population of
any country or place can only happen in two possible ways; through repro—
ductive change (natural increase) and/or net migration.

One of the most important demographic facts about a population
is its rate of change. How the population of a locality, a region or a state
changed to be what it is today in comparison with those of other areas
and how its population may be at some future date is one of the major
concerns of geographic analysis.

The present study focuses its attention on population change in
Central South—Mamprusi, Northern Ghana, during the twentieth century.

The analysis centers on the components of population change with emphasis
on the mechanisms by which the change took place.

For the early period population change was accounted for at the
local council aggregate level, that is, considering Central South-Mamprusi
together. For the period between 1948 and 1960 the enumeration areas of
the 1960 Ghana census were utilized in an attempt to account for differential
population growth patterns. Simple and multiple regression analyses were
the basic statistical techniques employed in estimating the relationships

between pOpulation growth and thirteen selected independent variables which

John S. Nabila
are directly or indirectly related to population change.

Central South-Mamprusi forms part of the Voltaic family and lin-
quistically belongs to the Mole-Dagbani association which covers a major
part of Northern Ghana and Upper Volta. The study area is known to have
been settled for a very long time, perhaps before the neolithic period.

One of the major aspects of the area's human geography is its
peopling by small groups from further North, beginning about the 15th century.
These immigrants from the North developed a system of traditional authority
and today there is a hierarchy of chiefdoms. This system of authority
has been of paramount importance in considerations of the general develop-
ment of the area, including trends in population growth and distribution.

Though earlier censuses were not very accurate, they do indicate
that population has been increasing in South-Mamprusi since 1900. At the
turn of this century both birth rates and death rates were very high. The
occurrences of epidemics, such as the influenza epidemic in 1918, cerebro
spinal menengitis in 1921 and sleeping sickness in the 1930s, raised death
rates. However, the establishment of health centers coupled with the
control and eradication of the epidemics has helped in considerably reducing
mortality rates, especially infant mortality.

There has been a general desire for children to help in the sub—
sistence economy, hence until quite recently children were regarded as an
economic asset. This factor, for a very long time, tended to raise fertility
levels. Indeed, the customs and religion of the people do favor high
fertility. Thus, though the death rates were high, the population continued
to increase considerably. Between 1911 and 1921 it increased by 54.3

per cent while between 1931 and 1948 it increased by 34 per cent. The reduced

John S. Nabila
growth rate in this latter period was caused by the sleeping sickness
epidemic. Between 1948 and 1960 there was an increase in population of.
about 54 per cent, thus giving an intercensal growth rate of about 3 per
cent per year.

In an attempt to understand the factors that account for population
change, it was hypothesized that traditional authority, density of population,
education and urbanization, have direct relationships with population
change while distance from main roads and urban centers have inverse relation-
ships with population change. The results of the multiple regression model
supported the hypotheses formulated at the beginning of the study. Cultural
considerations, namely traditional authority and related subjects are the
most important factors directly associated with population change. Den-
sity of population, urbanization, internal migration, and immigration from
neighboring African countries were also found to be important factors
affecting population growth. Long distance emigration from SoutheMamprusi
was found to be insignificant.

It was, however, difficult to explain from the results of the
multiple regression model or the data available to the author, the specific
contributions of the major components of population change, namely,
reproductive change and net migration. This was largely due to lack of

any relevant data on the vital statistics of SouthéMamprusi.

POPULATION CHANGE IN CENTRAL SOUTH-
MAMPRUSI, NORTHERN GHANA

By

John Sebiyam Nabila

A THESIS

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

MASTER OF ARTS

Department of Geography

1970

PREFACE

The author takes this opportunity to express his gratitude to

the Paramount Chief, Divisional Chiefs —- his father being one of them —-
and people of Central South—Mamprusi for their cooperation during his field
work.

Appreciation is also expressed to Dr. K. B. Dickson, Dr. George
Benneh and Mr. S. K. Gaisie, all of the University of Ghana for the initial
help they gave in directing the author's field research.

He is especially grateful to Dr. John M. Hunter, Department of
Geography, Michigan State University, for his encouragement and suggestions
in the writing of this thesis. In addition, the author is grateful to
Dr. Robert N. Thomas and Dr. Stanley Brunn all of Michigan State University,
Department of Geography, for their helpful suggestions.

The writer is especially indebted to his father, Wulugunaba—
Sebiyam, for his good understanding in helping the author to purchase a
Japanese Honda 125cc. (motor cycle) for his field work.

Finally,the author is grateful to the Scholarship Committee of
the University of Ghana for awarding him a post—graduate scholarship to

do further studies in the Department of Geography, Michigan State University.

ii

PREFACE.

LIST OF TABLES

LIST OF FIGURES.

CHAPTER

I

II

III

INTRODUCTION:

TABLE OF CONTENTS

Definition of Population Change . . .
Theories of Population Gorwth . . . .
Content of Population Geography . . .
Purpose and Scope of the Study. . . .
The Study Area. . . . . . . .
Content of the Study. . . . .

A REVIEW OF LITERATURE. . . . . .

UOUU>

E.
F.

THE PROBLEM AND STUDY AREA .

General Concepts of Population Growth
Population Growth in Ghana. .
Components of Population Change . . .
A Consideration of Variables.

1.

GJNChUIbLON

Basic Assumptions about the Study:
Dependent and Independent Variables .

Fertility Ratio . . . . .
Density of Population . .
Migration and Sex Ratio .
Place of Birth. . . . . .
Education . . . . . . . .
Traditional Authority . .
Urbanization and Distance
The Dependency Ratio. . .

DATA AND METHODOLOGY. . . . . . .

A.

Sources of Data . . . . . . .

1.
2.
3.

National Censuses . . . .
National Archives . . . .
Sample Survey . . . . . .

Ho
Ho
H.

Factor

Hypotheses.

Page
ii

vi

27

27
27
28
28

CHAPTER Page

B. Methods for Analysis of Problem . . . . . . . . . . 30
1. 1900- 1948 . . . . . . . . . . . . . . . . . . . 30
2. 1948—1960 . . . . . . . . . . . . . . . . . . 31
3. The Multiple Regression Analytic Method . . . . 31
4 Residuals from the Regression Model . . . . . . 35

IV POPULATION GROWTH 1900-1948. . . . . . . . . . . . . . . . 37

A. PeOpling of the Area. . . . . . . . . . . . . . . . 37
B. Population Growth 1900—1948 . . . . . . . . . . . . 42
1. Population Growth 190091921 . . . . . . . . . 43
2. Population Change Between 1921 and 1948 . . . . 49

V POPUIAATION GIANGE "‘ 1948—1969. 0 o o o o o o o o o o o o o 59

A. Population Change 1948-1960 . . . . . . . . . . . 59
1. Results of the Multiple Regession Analysis. . . 64
2. Factors Accounting for Population Growth.. . . . 65
(a) Traditional Authority. . . . . . . . . . 68
(b) Urbanization . . . . . . . . . . . . . . 74
(c) Education. . . . . . . . . . . . . . . . 75
(d) Another Locality . . . . . . . . . . . . 76
(e) Density of Population. . . . . . . . . . 78
(f) Foreign Born . . . . . . . . . . . . . . 78
(g) Sex Ratio. . . . . . . . . . . . . . . . 79
(h) The Distance Variables . . . . . . . . . 81
(i) Born in this Locality. . . . . . . . . . 82
(j) Dependency and Fertility Ratios. . . . . 85
(k) Born in Another Region . . . . . . . . . 86
3. A Consideration of the Residuals. . . . . . . . 86
Post-Censal Period - 1960—1969. . . . . . . . . . . 92
1. Fertility and Mortality . . . . . . . . . . . . 92
2. Internal Migration. . . . . . . . . . . . . . . 96
3. Immigration . . . . . . . . . . . . . . . . . . 96
4. Out-migration . . . . . . . . . . . . . . . . . 99

VI SW AND CONCLUS IONS 0 O O O O O O O O O O O O O O I O O 102
GLOSSARY O O O O O O O O O C O O O O O O O O O O O O O O O 110
APPENDICES . O O O O O C I O O O O O O O O O O O O O O I O 111

A(l) Dependent and Independent Variables . . . . . . . 111
A(2) Matrix of Simple Correlations . . . . . . . . . . 112
B. Enumeration Areas with Per Cent Population Change

and Residuals . . . . . . . . . . . . . . . . . . 113
C. Questionnaires for Field Survey . . . . . . . . . . 114

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . 116

iv

TABLE

10.

ll.

12.

l3.

14.

15.

16.

17.

LIST OF TABLES

Hierarchy of Traditional Authority . . . . . . . . . . .
Composition of Population by End of 19th Century . . . .
Population Growth — 1900-1921. . . . . . . . . . . . . .
Ethnic Composition of Population in Gambaga 1921 . . . .
Population Growth 1921—1948. . . . . . . . . . . . . . .
Incidence of Sleeping Sickness in Ghana 1932-1939. . . .
Immigrants from North—Mamprusi and Upper Volta . . . . .

Number of People Enumerated from Northern Territories in
the Colony and Ashanti, 1948 . . . . . . . . . . . . .

Population in Mamprusi, 1960 by Divisions. . . . . . . .
Composition of Population by Ethnic Groups, 1960 . . . .
Analysis of Variance Test for Significance of the Simple

Correlations and (b) the Standard Partial Regression
coeffiC1ents O I O O O O O O O O O O O O O O O O O O 0

Summary Table for the Stepwise Regression Model (LSADD and

LSDEL) O O O O O O O O O O C O O I O O O I C O C C O 0
Ghana, Total Fertility by Ethnic Group and Urban/Rural .

Short and Long Range Mobility by Ethnic Group and Sex,
Northern Ghana . . . . . . . . . . . . . . . . . . .

Mean Number of Children Ever Born to WOmen by Age Group.

Total Fertility and the Number of Deaths for the Past
NO Years 0 O O O O O O O O O O O O O O O O O O O O O 0

Long Distance Migration in SouthuMamprusi. . . . . . . .

Page
42
43
44
48
51
55

57

58
61

62

66

67

70

71

93

94

100

FIGURE

10.

ll.

12.

13.

14.

15.

LIST OF FIGURES

Central South-Mamprusi-Location. . . . . . . . . . . . .
Central South-Mamprusi-Relief. . . . . . . . . . . . . .
Central South-Mamprusi-Cross-Section . . . . . . . . . .
Central South-Mamprusi 1960 Census Enumeration Areas .

Boundaries of 1960 Census and Author's 1969 Field Survey
Enumeration Areas for the Multiple Regression Analysis .
Evolution and Distribution of Settlements. . . . . . . .
Ghana, Density of Population - 1921. . . . . . . . . . .
Incidence of Sleeping Sickness in Ghana 1919-1939. . . .
Northern Ghana, Density of Population, 1960. . . . . .

Population Change 1948-1960. . . . . . . . . . . . . . .
Population Pyramid - 1960. . . . . . . . . . . . . . . .
Distribution of Population - 1948. . . . . . . . . . . .
Distribution of Population - 1960. . . . . . . . . . .

Residuals from the Step-wise Deletion Program. . . . .

Page

29
32
33
40
47
53
6O
63
80
83
84

88

CHAPTER I

INTRODUCTION: THE PROBLEM AND STUDY AREA

Definition of Population Change

 

Population change is a dynamic equilibrium between forces of
increment and forces of decrement. Thus, change in the population
of any country or place can only happen in two possible ways: through
reproductive change (often called natural increase)1 and net migration.
Reproductive change is the balance between births and deaths while
net migration is the balance between in-migration and outdmigration.

AccOrding to Bogue, if we know the population count as of a
particular date and keep track of the number of occurrences of each of
these four components of growth, we can calculate the population of any
later date. By using the demographic book-keeping equation, this can
easily be worked out, for example:

Pt - Po + (B-D) - (MieMo)
Where Pt is the later population count, P0 is the earlier popu-

lation count; B= births during interval between Pt and Po; D - deaths

 

1Reproductive change is not a more "natural" source of change in
comparison with births and deaths. Besides the net effect of fertility
and mortality is not always to increase population. Hence, the term,
"natural increase," is a complete misnomer: Bogue, Donald J., Components

 

of Population Change, 1940-50; Estimates of Net Migrations and Natural
Increase for Each Standard Metropolitan Area and State Economic Area.
Scripps Foundation, 1957, p. 1.

2
deaths between the two periods; Mi and Mo = in-migration and out-migration

respectively, for the interval between Pt and Po.

Theories of Population Growth

 

Theories of population have been suggested by many people in an
attempt to explain the main influences, especially with respect to fer-
tility, of population growth. Coontz has classified these theories into
three categories: biological, cultural and economic.3

Biological theories stress that the law regulating human popu-
lation growth is basically the same as that regulating the growth of plants
and animals. This view which relates fertility to nutrition and density
has found different exponents in the last two centuries. De Castro,
for instance, hypothesized that there is an inverse relationship between
protein intake and human fertility. In his analysis, he associated low
productivity and high productivity to hunger, especially specific hunger
of proteins, minerals or vitamins.

Cultural theories of population growth emphasize the role of man's
character and culture in influencing his fertility. Declining fertilities
in advanced countries is said to be due to the influence of desires for
higher social and economic status and for more luxuries and pleasures in

life. More often than not, many people have suggested increased education

 

2Bogue, Donald J., Principles of Demography. John Wiley and Sons,

 

1969.

3Coontz, S. E., Population Theories and their Economic Interpreta-
tion, Routledge and Paul, London 1957.

4de Castro, Josu, Geografia da fome, (Geography of Hunger).
Livraria - Editora da Casa do Estudiante to Brazil, Rio de Janeiro, 1953.

 

3

and urbanization as determinants of fertility decline in some parts of
the world, especially in the advanced nations.

Economic theories of population growth emphasize the significance
of economic relations in social change. It has often been suggested that
there is an inverse relationship between social status and fertility. Con-
ditions of fertility, migration and geographical distribution of population
are regarded as being essentially responses to economic conditions. Stress
ing the role of economic factors, Zelinsky suggests that the economic char-
acteristics of an area exert a much more direct effect upon its population

than do its physical characteristics.5

Content of Population Geography

 

Population geography has been described as the empirical, spatial
and mathematical study of human populations. It focuses its attention
areally on three readily observable human phenomena: (1) change in popu-
lation size (growth or decline) (2) the composition of the population and
(3) the distribution of population in space. It is interested in how the
size, composition and distribution of population vary throughout space,
whether at present or through time. Thus, population geography is con-
cerned with demonstrating how spatial variations in change, composition and
distribution of populations are related to spatial variations in the nature
of places. A place in this context, may be a territory of any extent from

a minute area to the entire surface of the earth.6

 

5Zelinsky, Wilbur, A Prologue to Population Geography, Prentice-
Hall, Inc., New Jersey (1966), p. 34.

 

6Thomas, Robert N., Seminar notes on population geography, Geo.
804.

4

Therefore, the study of population change is a legitimate field
for geographers. How the population of a locality, a region, a state, or
even a continent, changed to be what it is today in comparison with those
of other areas, and how its population may be at some future date, is one

of the major concerns of geographic analysis. Trewartha has remarked that,

"....any comprehensive geographical analysis of a region should
take into consideration, factors of differential population growth
rates, both as it applies to areas beyond, and outside the re-
gion in question, and likewise, to the several localities which
comprise the region. Both inter-regional and intra-regional
contrasts in differential rates of growth carry in their train
economic and political and social consequences which the student
of social and cultural phenomena may not ignore."7

Purpose and Scope

 

This is a study of the change of population in Central South-
Mamprusi, Northern Ghana. The study focuses on the changing patterns of
population growth, especially in the 20th Century. An attempt is made to
account for the factors which affect population change (growth or decline)
in South-Mamprusi. In this way, the author gives an account of differential
growth rates in the various Census Enumeration Areas which constitute the
study area.

In the analysis, the study will try to answer the following

questions:

1. What are the major components which are responsible for
population change?

2. Which are the areas with high population growth rates
and why?

 

7Trewartha, Glen T., "A Case for Population Geography," Annals
of the Association of American Geographers, Vol. 43, No. 2, 1953, pp.
71-790

5

3. Which are the areas with declining population change
rates and why?

4. What is the role of the traditional system of authority
in population change?

5. What relationship exists between population changes in
the various enumeration areas and the major economic
activities carried out by the people?

6. What have been the conditions in population change
during the post-1960 population census period?

The StudypArea

 

Central South-Mamprusi has an area of 1,692 square miles and is
56 per cent of the whole of South-Mamprusi, which has an area of 2,996
square miles (Figure 1). Central South-Maprusi lies approximately between
100 North and 100 40' North. Its boundaries are clearly defined by the
White Volta-River Morago, River Nasia and the former Anglo-German boundary.

For the most part, the area lies between 400 ft. and 1,000 ft.
above sea level, but rises to about 1,500 ft. in the eastern part of the
area. Indeed, it is the dip slope of the Gambaga escarpment (Figures 2
and 3). Apart from a small area around Kpasenkpe, which overlies granite,
the whole district lies on the Voltaian geological systems consisting of
sandstones and shales. Light sandy soils are common and there are some
silts and clay in the valleys.

The climate is Sudanese-type with a pronounced dry season, i.e.,
when there are In) rains, during the period from November to March and
a wet season which begins in March-April, and continues to October with a
peak in August-September. In the dry season,from November to February,
the "Harmattan" - a dry desiccating wind - blows more or less continuously
from the north-east. The mean annual rainfall for the area is 45 inches.

The natural vegetation is tree-savannah i.e., with.ta11 grass and rather

6

LOCATION CENTRAL SOUTH “MAMPRUSI (SHADE?)

 

 

 

 

 

 

 

 

 

 

 

Ho). -- AAZO “'1 . —-.M~ |° o
T I T Inwku I "
U P p E R oBol~oton~t
_ %
R E G l O N
| 0W0 ——IO°
"J
V ’TT-
T3 ’
> .—
g R
O
9.__ Damongo —‘9°
.- / ”T
\~ _
s°— F80
a R o N A H A F o ‘
\ (9
4
r~ ’
Sonyani 0 \\“"’ €429. \_\ \ c:
1 I ’
79 'fi—- ’a- It 55 '4 4A TD" I ‘1 -—‘7°
\\ I/ .Kumasi I ( / 3.40
19— \‘Ir [EASTERN '
J l R s c I o N,’ 0
6‘ 7x J
‘5‘ x
\ -—1 be
~50
1 1 1 {azurer
3° 2° l° 0° |°

 

 

Figure l

 

 

 

 

.H .m .woma .commA «wmsumwmzlnusom Hmuusuo mo >nmmwmomu Hmuwuoumwm a< .m snow .maflnmz umuw<
N shaman

 

 

 

2.0 . bro I 9.0 N/ ._
890. , . l

          

JOOoo.
poo. .90... :20 5v
. .30. 3.0! :93 In
690. .20.. 39.0 2N
woo. 3.0! so? 2.

 

 

 

 

 

 

 

 

.90 9.0 9.0 ._

 

mO<Z.<mO OZ< “mama
ESE—.32 IpDOm 45:sz

 

.H .a .wooa .mmDHMEmZILummm Hmuucuquo afimmawomu Hmoauoumum :< .m snow .mawnmz pmuw<

 

 

 

 

 

 

m shaman
3k 8a1-l &O. and ”‘33, 31‘qu M m q M. n . O d‘x‘}!
00... l \\ \IMFI\ \Aoom
II.gT....¢: IIIIIIIII \u I I480.
000. jI ...z..uI I. I .33... ill. .2... or“;

1».
08_.I||I|..I|IIIIIII.IIIIIIII}! &.III4003
OOON lllllllllllllllllll (II 1 lllllllll LfiOOON

m 2

9:6 020$ azuzafiomw .0326 e: o 20:03 306 <

9
low trees. However, the vegetation has changed as a result of the
activities of man, for example, farming and annual burning of the bush in
the dry season. In fact, in almost all places, except along rivers, it
is now a biotic climax vegetation of low grass and scattered trees.

Quite possibly the area was settled for a very long time, per-
haps before the neolithic times. Starting from about the 15th Century it
has experienced in—migration, especially from the north. Before the turn
of this century the people acquired their wealth from entrepst trade
between the Western Sudanese states and Southern Ghana. This was sup—
ported by subsistence agriculture. When the trade dwindled at the begin-
ning of the twentieth Century, especially after Bawku and Bolgatanga
supplanted Gambaga and Walewale as commercial centers, the economy changed
to one almost solely based on subsistence agriculture. Today, many
agricultural innovations are being introduced, such as irrigation, the
use of fertilizers, the use of farm machinery and the cultivation of
cash crops like rice, tobacco, groundnuts and cotton.

The social and economic development of SouthéMamprusi, just as
other parts of Northern Ghana has lagged behind that of other areas in
Southern Ghana. Hence, today there are not any significant urban centers.
Indeed, the largest settlement, Walewale, had 4,493 in 1960; thus, it
failed to qualify as an urban center according to the definition of the
1960 Population Census.8 Within South-Mamprusi itself, or even considering
its space relations with its immediate neighbors, the few rural-urban

centers do render important commercial, medical, educational and other

 

8The Ghana 1960 Census Reports defines an urban center as any

-settlement with a population of 5,000 people.

10

economic and social functions to the rural areas.

There were two major considerations in the choice of this study.
In the first place, very little has been written on this area. The only
writings on it are general works on Ghana as a whole,1iterature on only
Northern Ghana and a few diaries of the former Colonial District Commis-
sioners which primarily describe the routine activities of the officers.
The author wrote a B.A. thesis on the historical geography of the area
and the present study is a follow-up. It is hoped, by focusing on the
human populace, especially how they have increased, that the two studies
will, by and large, help in future related studies of the area. Secondly,
the author comes from the area, a factor which made his field research

relatively manageable within the resources and time available to him.

Content of the Study

 

Chapter II has been devoted to a review of literature on popu-
lation change and related subjects. It is hoped that this will help to
bring to light some of the factors that may account for population change
in South-Mamprusi. In addition it should help in directing the major
analysis of this study in terms of the hypotheses to be formulated as well
as variables to be used for the study. Chapter III gives an account of
the sources of data and methods used for the analysis. Chapter IV focuses
attention on population change from 1900 to 1948, with an introduction
on the peopling of the area; Chapter V gives an account of population
change to the present. Finally, Chapter VI contains the summary and
conclusions (findings and implications) of the study. A glossary has been
provided at the end of the analysis to help the reader understand a few

Mamprusi words used in the main text.

CHAPTER II

A REVIEW OF LITERATURE

A. General Concepts of Population Growth

 

One of the most important single demographic facts of a popu-
lation is its rate of growth. For the rate at which a population grows
affects not only its size and numerical increase but also its composition.
Population growth is both cause and consequence of social and economic
conditions in any society and is, therefore, a subject worthy of study
that cannot be ignored. For a very long time the concept of population
growth has received attention from many disciplines. Some of the known
demographic theories developed in respect of this single factor about
human populations have already been mentioned in Chapter I. Here a con-
sideration is given to the important general ideas pertaining to popu-
lation growth.

A few of the general concepts about the rate of population growth
are discussed below. (a) The Malthusian theory postulates that population
increases in a geometric ratio while subsistence increases in an arithmetic
ratio. History has shown that Malthusian theory lacks the quality of

universality and his critiques have often noted the factor of technology

 

1Malthus, Thomas, Essays on the Principle of Population. 1789.

 

London.

11

12
in food production.2 However, his supporters point to the fact that food
production could have a limit and the rapid growth of population could be
a threat to the very existence of man. (b) The theory of logistic curve
of population growth, formulated by Verhult in the first half of the
nineteenth century but now usually associated with R. Pearl and G. V. Tule
(1871) or R. Pear, and L. J. Reed (1920), states that growth is a function
of the size and density of population itself. That is, with an increase in
the size and density of a given population living in a constant environment,
the birth rate will decrease and the death rate will increase, hence the
rate of growth decreases. In addition, population size is limited by
cultural conditions and technology of production.3 This theory, just as
the Malthusian theory, is under constant attack because of its deterministic
flavor. Indeed, it fails to take into consideration possible changes in the
technology in food production, level and changes of aspirations of mankind
and changes in human reproductive patterns. (c) Ibn Khaldun and E. F.
Wagemann feel that population growth is cyclical. To Khaldun, population
change is a consequence of economic, political and socio-psychological
conditions. He held that besides the actual situation expectation of men
about productivity in the future played an important role in changing size

of population. To Wagemann, improvements in technology of production would

 

2Wrong, Denis, Population and Society. New York, Random House,
1962;

Bogue, Donald J., Principles of Demography. John Wiley & Sons,

 

1969.

3Spengler, Joseph J., and Otis Dudley Ducan (eds.), Population
Theory and Policy. Illinois, The Free Press, Glencoe, 1956, pp. 30-32;
Pearl and Yule: "The Growth of Population and the Factors Which Control
It." The Journal of the Royal Statistical Society, LXXXVIII (1925), p. 38.

 

13

help the society to overcome the law of diminishing returns and thus
enable population to grow larger. They all hold that at a certain point
the advanced socio-economic conditions so created will result in de-
population, which will in due course be replaced by overpopulation.4

(d) All the world's countries, as they develop and become modernized,
pass through a series of changes in birth and death rates and, accordingly,
in population growth rates.5 This is the basic concept underlying the
popular Demographic Transition Theory. It has a notion of a demographic
development correlated with stages, from Pre-Industrial (Stage 1) to Pre-
Western (Stage II) and finally to Modern Western (Stage III). Stage I
is represented by the stability of population size caused by the balance
of high, though fluctuating, death rates with high birth rates. During
Stage II, population begins to experience the effect of industrialization
and modernization. Improvement in nutrition and health standards helps
bring the mortality rate down, while fertility rates remain high. The
gap between mortality and fertility is widened, resulting in the rapid
growth of population. Most of the emerging countries of Asia, Africa and
Latin America are experiencing this imbalance at present. At the last
stage, associated with advanced industrialization, preference for small
families increases and consequently the fertility rate declines. Thus, a
balance between fertility and mortality is again approached and the growth

of population slows. Only the advanced industrialized countries have

 

4Spengler and Duncan, op. cit., p. 8.

STrewartha, Glenn T., "A geography of population Characteristics,"
in Geography of Population, Paul F. Griffin (ed.), Fedron Publisher, 1970,
pp. 39-64.

14
reached this stage.6 The three stages are sometimes labeled, with respect
to the change in population size only, as High Growth Potential, Transi-
tional Growth and Incipient Decline.

Many people have questioned the universality of this transition
theory, especially whether it is applicable in the deve10ping countries
since it was developed after the pattern of population growth in Europe.

The assumption was made that the contemporary developing world will follow
the pattern which occurred in Europe in their process towards modernization.
Past trends of population growth in many places do not conform to the theory,
while those which conform to the model show variations in the timing and
duration of successive stages of evolution. Also, the demographic and
socio—economic conditions as postulated by the theory during Stage I were
much different in the Asian and African countries in comparison with those

in Europe.

Despite these shortcomings, the transition theory is an "ideal
model" for scientific studies in various fields. The core of the model,
despite the variations in real phenomena, is the change from balancing
high mortality rate with high fertility to balancing low mortality with low

fertility. This concept will be utilized for this study.

B. Population Growth in Ghana

 

Population growth in Ghana has been rapid, even by the standards

of most of the developing countries. Between 1921 and 1960, Ghana as a

 

6Wrong, Dennis, op. cit., 1962;

Bogue, Donald J., op. cit., 1969;

Coale, Ansley J., "How a population ages or grows younger,"
in Ronald Freedman, (ed.), Population: The Vital Revolution, New York,
Doubleday and Co., Inc., 1964.

 

15

whole trebeled its population and between 1948 and 1960 the national
growth was 63 per cent, that is, an intercensal growth rate of about 4.2
per cent per year.

Thus, between 1948 and 1960, of the 69 Local Councils of Ghana
(1960) there were 21 councils with vigorous population growth rates of
80 per cent increase and over, e.g., Sunyani, 115%; Tema, l,400%;

Accra, 166%; Kumasi, 166% of the rest, 25 councils had an average growth
of 50-79 per cent increase while 23 councils had slack growth or absolute
decline of 49 per cent increase and below.8 Another striking feature is
the growth in size and number of urban centers. During the intercensal
period the urban rate of growth was more than double the rural rate -—
116 per cent and 52 per cent or 6.6 per cent and 3.6 per cent per annum
respectively -- while the proportion of urban to total population rose
from 17 per cent to 23 per cent.

Central South-Mamprusi falls within the third national category
of population growth, that is, slack growth, with an intercensal growth
rate of 52 per cent between 1948 and 1960. An analysis of differential
growth rates in Mamprusi reveals that there are a few areas which fall
within the first category. For example, Walewale, Langbinsi, Nagbo, and

Burugu have growth rates of 139%, 127%, 124%, and 109% respectively;

 

7Caldwell, John 0., "Population Change," in Walter Birmingham,
I. Heustadt and E. N. Omaboe (eds.), A Study of Contemporary Ghana,
Vol. II. Some Aspects of the Social Structure, London, 1967, p. 78.

 

8Hunter, John M., "Regional patterns of population growth in
Ghana 1948-69," in Essays in Geography for Austin Miller, J. B. Whittow
and P. D. Wood (eds.), 1965, pp. 272-90.

16

while Zangu, Fio and 800 Daboya have growth rates of -4%, -7%, and

-l4% respectively.

C. Components of Change

 

An analysis of population change often involves an examination
of the immediate components of change, namely, births, deaths and in—
and out-migration. Each of the components is usually measured as a rate
and the commonest are the "crude" rates: crude birth, death and migration
rates. A crude rate is simply the number of events of a given type that
occur in a year divided by the mid-year population. Usually this ratio
is multiplied by 1000 to show the number of events per 1000 people. In

algebraic terms, the crude birth and death rates are:

ytotal births
midyear population

 

CBR X 1000

_ total deaths
CDR _ midyear population X 1000

 

The crude rate of reproductive change is the balance between the crude
birth rate and the crude death rate. In like manner, net migration can
be obtained through the same process, thereby enabling us to arrive at
the growth rate, that is, reproductive change plus net migration.9

The crude rates, though the simplest and commonest for population
change analysis, are not adequate measures of change since they are cal-
culated, especially in respect of births and deaths, without regard to sex

and age composition of the population. The factor of sex and age composition

 

9Bogue, Donald J., op. cit., 1969;
Kuczyski, Robert, The Measurement of Population Growth. New York,

 

1936'
’ Baclay, George, Techniques of Population Analysis. London and

New York, 1958.

 

17

of the population. The factor of sex and age composition is eliminated by
relating births and deaths to specific female and age groups. In this way,
various fertility rates, for instance, are now in constant use for demographic
analysis. A few of these are general fertility rate, age—standardized fer—
tility rates, age—specific fertility rates, gross reproductive and net re-
productive rates.10

From the above, it is quite evident that the availability of data
on the components of change is of paramount importance for a sound analysis
of population change. Such data can be obtained from vital statistics -~
registers of births, deaths and migration, ---- and Census reports. Detail-
ed demography of Ghana is not yet known. There is no information on fer-
tility or mortality representative of the various sections of the country.
Birth and death statistics are published regularly for a few of the major
cities and towns, but they can hardly be held representative of vital rates
in the country. Registration of births and deaths in many parts of Ghana,

Mamprusi inclusive, is nonexistent. Indeed, most of the studies on the

demography of Ghana have been based on either general reports of a few

 

10Bogue, D. J. and Palmore, James, "Some empirical and analytical

relations among demographic fertility measures with regression models
for fertility estimation," Demography, Vol. 1, 1964, pp. 316-338;

Bogue, D. J. (1969), op. cit.;

Kyczynski, op. cit., 1936;

Ryder, Normal B., "Fertility," in Philip M. Hansen and O. D.
Duncan (eds.) The Study of Population: An Inventory and Appraisal,
University of Chicago Press, 1959.

18

sample surveys.11

The first population count to be carried out in Ghana took place
in 1891. However, this was restricted to the southern part of Ghana.
After that count, there were subsequent censuses every ten years (except
in 1941 due to the 2nd World War). A count was taken in 1948 and then
in 1960; and the latest was made in March—April, 1970. Apart from the
1960 census, the accuracy of the first censuses is greatly in doubt. Even
with the 1960 census there are no direct data on births, deaths and migra-
tion for specific periods. According to Caldwell, since Ghana is not
covered by a birth and death registration system, an estimate of the rate
of population growth cannot be made from the results of a single census
combined with the known vital rates of the time. It must depend on two
censuses or an examination of the age structure depicted by the most recent
census.12

A population change is said to be explained when it has been sub-
divided into its major components and subcomponents. Therefore, this study
involves the analysis of the components with emphasis on the mechanisms
by which the change took place. These mechanisms will involve an examination

of variables that are directly or indirectly related to population change.

 

11Caldwell, J. C., 1967, op. cit.;

, "Fertility differentials as evidence of incipient
fertility decline in a developing country - the case of Ghana," Population
Studies, Vol. XXI, No. 1, July, 1957;
, "Fertility attitudes in three economically con-
trasting rural regions of Ghana," Economic Development and Cultural Change,
Vol. XV, pp. 217-38;
Gaisie, S. K., Dynamics of Population Growth in Ghana. Legon, 1969;
Pool, D. T., "The Ghana Fertility Survey," paper presented to the
first African Population Conference," Ibadan, January, 1966.

12Caldwell, 1967, op. cit., p. 78.

 

 

 

19

D. A Consideration of Variables

 

1. Fertility Ratio

 

One crude measure of fertility is the fertility ratio (child/
woman ratio). This is computed by dividing the number of persons under
5 years old in the population by the number of women 15-49 years of age.
This is often regarded as the effective fertility or the fertility re-
maining after the bulk of infant mortality has occurred.13 However,
Bogue and Palmore have indicated that the ratio underestimates fertility,
for although the deaths of women in the child bearing years tend to compen-
sate for deaths to children, the net result is that the ratio is biased.14
In any case, with the absence of vital statistics, this is quite a sound
measure of fertility. In Ghana many women get married by 20. Even though
early marriages may adversely affect the woman's health, it is a factor
which considerably increases fertility, both because it prolongs the actual
chi1d~bearing period and because it is the young female age-group that has
the highest fertility. In Ghana, the 1960 Census Report gives data for

the age group 15-44 years. Besides, as Caldwell has indicated, this is the

most effective fertile age group.15

2. Density of Population

 

Density of population was suggested by Bogue and Harris as being
potentially important in explaining differences in growth between metropoli-

tan areas in the United State of America. To them, dense settlement and

 

13Gabil, Wilson N. and Cho, Lee Jay: "Methodology for the measure-

ment of current fertility from population data on young children,"
Demography, Vol. II, 1965, pp. 50—68.

 

14Bogue and Palmore, 1965, op. cit.

15Caldwell, 1967, op. cit., pp. 87 and 102.

20
intensive use of land may facilitate growth in a Standard Metropolitan
Area by providing favorable cost and market conditions for new and
expanded industry, i.e., making room for more people.16 In like manner,
Thomas considered density as one of the factors areally associated with
population growth. In Chicago, he found a negative association between
density and population growth and suggests that cities with "large popu—

lation increases would be those with lower densities,"

will move into these noncongested cities.17 In Studies in Demography of

since many people

 

Pakistan, 1967, edited by Warren C. Robinson, density of population was
found to have significant statistical correlation with fertility and
population change.

In answering the question, "At what stage, and to what extent
will pressure on the land act as a determinant of population growth?",
Hunter found out that density gives a poor correlation with growth of
population in Ghana.18 On the other hand, Caldwell indicates that the
densely populated Southern Ghana has higher growth rates than sparsely

populated Northern Ghana.19

3. Migration and Sex Ratio

 

It is predominantly the young adult who usually migrates.

In Ghana, contrary to Ravenstein's theory, men are more migratory than

 

16Bogue, Donald J. and Dorothy L. Harris, Comparative Population and
Urban Research Via Multiple Regression and Co-variance Analysis. Scripps
Foundation, 1954.

17

 

Thomas, Edwin H., "Areal associations between population growth
and selected factors in the Chicago urbanized area," Economic Geography,
Vol. 36, 1960, pp. 158-170.

 

18Hunter, 1965, op. cit., pp. 272—90.

lgCaldwell, 1967, op. cit., pp. 107—109.

21
women.20 From the age and sex factors, it has been found that many pull
areas have high sex ratios within the 15—44 years age group, while push
areas have low sex ratios. Hunter found significant correlation between
sex ratio and growth of population: r=0.686; and r=0.789 between immi-

gration and sex ratio.

4. Place of Birth

 

During the 1960 Census, persons were classified by place of birth,
according to whether they were born in (1) the locality of enumeration
(2) another locality in the same administrative region (3) another ad—
ministrative region or (4) abroad, including Africa. A classification of
the population of a given area by place of birth makes possible a crude
measure of the volume and sources of migration into the area during an
indefinite number of years. An analysis of migration based on this is very
crude because such data does not take into consideration migrants who have
died between the time of migration and the date of the census; of those
having returned to their places of birth or moved on to other areas, after
migrating to the area in question, or of previous migrations between the
time of birth and the time of entering the given area. However, it is
possible to obtain meaningful general patterns of migration from the type
of classification used in Ghana. Numerous studies on the demography of
Ghana have used this variables for analyzing both migration and population

growth.21

 

0Ravenstein, E. G., "The laws of migration," Journal of the Royal

Statistical Society, Vol. 48, (1885), pp. 167-235; Vol. 52 (1889), pp.

21Caldwell, 1967, op.cit., pp. 111-146;

Hunter, 1965, op. cit., pp. 272-90.

22

5. Education

A survey in Ghana revealed that of males and females who have
never been to school, 27% and 16% respectively migrated at some time to
an urban area; but 67% and 61% of those with high school or university
education had done 30.22 It has been found that changes in the traditional
family system, introduction of western innovations and the like are more
prevalent among the educated and Urban elite than it is with the rural

and the uneducated population.23

6. Traditional Authority

 

In addition to modern forms of administration, many places in
Ghana have a system of chieftaincies. These have been in existence for
a very long time, a few of them from time immemorial. Centers of
chieftaincies have often served as points for the grouping of people

together and dense populations were associated with stable chieftaincies.24

 

22Caldwell, John C., "The Determinants of Rural-Urban Migration
in Ghana." Population Studies, Vol. XXII No. 3, pp. 361-77.

23Caldwell, John C., Population Growth and Family Change in Africa.
The New Urban Elite in Ghana. Canberra, 1968;

Jahoda, Gustav, The Social Background of a West African Student
Population (Part I and II) British Journal of Sociology, Vol. V, pp. 355-
69, and Vol. VI, pp. 71-9;

Omari, T. Peter, "Changing Attitudes of Students in West African
Society Towards Marriage and Family Relationships," British Journal of
Sociology, Vol. XI, pp. 197-210 (1960);

Busia, K. A., "Some Aspects of the Relation of Social Conditions
to Human Fertility in the Gold Coast, in Larimer, Culture and Human Fer-
tility, Paris (1954), pp. 341-50.

24Dickson, K. B., A Historical Geography of Ghana. Cambridge,
1969, p. 49.

 

 

 

 

 

23

In a study of South—Mamprusi, this latter factor was found to be of
paramount importance in the consideration of peopling and the density
of population.25 Most of these chieftaincies in SouthaMamprusi are
very rural and show some of the typical population characteristics of
rural Ghana, for example, extended and large families (polygamy).

Musham has shown that fertility of women living in polygamous
marriages is about twice that observed in monogamous marriages of the
same social and cultural milieu. The total number of children born in
a polygamous family (by all wives together) exceeds the total number of
children born in monogamous families. This is probably the main
demographic raison d'gtre of polygamy: to ensure a greater number of

offspring to a man.26

7. Urbanization and the Distance Factor

 

Urban centers as receiving points of population from rural areas
is an established phenomena in most of the countries of the developing
world. Apart from having concentrations of population, urban centers are
known to be destinations of innovations introduced in the developing world.
The role of the urban elite in Ghana in terms of growth of population,
family system, education, and the like, has been discussed by Caldwell in

his book, Population Growth and Family Change in Africa: The New Urban

 

 

 

Elite in Ghana.27 Caldwell found rural-urban differentials in fertility

 

25Nabila, John S., An Historical Geography of Central SoutheMamprusi

Thesis presented for the B.A. Degree. University of Ghana, Legon-Accra.
(Unpublished).

26Musham, W. V., "Fertility of Polygamous Marriages," Population
Studies, Vol. 10, Part 1, July 1956, pp. 2-16.

27Caldwe11, J. C., 1968, op. cit.

 

I'D- 5. . .,.I .¢-m|.b
I
.

24
and population growth in Ghana. Other studies, too numerous to mention
here, in other parts of the world have proven this factor of rural-urban
differentils in fertility, rate of population growth and other related
phenomena.

From these urban centers, innovations spread to rural areas and
therefore towns near them are at a better advantage than settlements that
are farther away. A major employment characteristic of an urban center
is the existence of people engaged in secondary, even tertiary, activities.
A few of these are usually carried out by migrants. The classification
of the employed population into (a) Employed in Agriculture and (b) Other
(in the 1960 Census) makes it possible for the "other" to be designated

as the population engaged in secondary or tertiary activities.

8. The Dppendency Ratio

The Dependency Ratio is the ratio between the total population
under 20 years, plus 65 years and over (the dependent population) on one
hand, and the total population aged 20 to 64 years. The ratio purports
to measure how many dependents each 100 persons in the productive years
must support. Bogue has pointed out that the ratio is directly a function
of the components of population growth.28 The minimum dependency load
in the long run is that obtained when the population is barely replacing
itself at the lowest possible combination of mortality and fertility.
Population growth always implies a higher dependency load than do stationary
numbers. This variable will, therefore, be used with the hope that with
the absence of vital statistics it could serve as a useful indicator of

v

28Bogue, Donald J., 1969, 0p. cit., pp. 155-156.

25

population growth differentials in the enumeration areas.

E. Basic Assumptions about the Study: Hypotheses

 

For the purpose of analyzing population growth in Central South-
Mamprusi, the following assumptions have been made with reference to the
major components of population change.

1. Reproductive change (natural increase) is the major source
of population change in the area under study.

2. There is a close relationship between traditional authority
and growth of population.

3. Men migrate for long distances more than women, consequently
the sex-ratio of the age group (15-44 years) is a good
measure of migration patterns. For it is mostly this age-
group which engages in long distance migration.

4. There is a direct relationship between density of popu—
lation and growth of population.

5. There is an inverse relationship between density of popu—
lation and distance from main roads. Thus, there is also
an inverse relationship between population growth and
roads.

6. The urban centers are areas of rapid population growth.

By the very nature of their functions there is an inverse
relationship between population growth and distance from
urban centers.

7. There is a close relationship between education and the tenden-
cy to migrate. There should therefore be a direct relationship
between education and population growth.

In order to understand the mechanisms for population change in the
area, the following factors related to the components of change either
directly or indirectly were used as variables for the analysis. Multiple
and simple regression are the basic statistical techniques employed in
estimating the relationships between population growth and the selected
factors. Population growth between 1948 and 1960 is the dependent variable

for the study, while the observations (unit areas) are based on the Enumer-

ation Areas (1960), and the authors field work in 1969.

26

F. Dependent Variable: Population Change (1948-60) - Variables.

 

. Sex Ratio
. Density of Population
. Per cent born in locality of enumeration

Independent Variables: l
2
3
4. Per cent born in another’ locality
5
6
7

 

Per cent born in another ration
Per cent born outside Ghana
. Population 15 years and over with Present

and Past Education.

8. Dependency Ratio

9. Fertility Ratio

10. Tradition Authority

11. Distance from nearest Urban Center

12. Level of Urbanization

13. Distance from Main Roads

Blalock has pointed out that the use of many variables, especially
where their implications are very obvious, is likely to obscure the results
of a regression analysis.29 There is a further problem in this study in
the use of variables pertaining to birth place. They are percentages and,
therefore, the use of all of them is likely to introduce multi-collinearity
since they add up to one hundred. Even though these shortcomings do exist,
in the absence of any reasonable data on components of growth, such as birth
rates, death rates and migration data, the inclusion of such variables will
help indicate some significant associations about the mechanisms of popu-
lation change. Indeed if there is a distortion in the total variation ex-
plained (R2) the simple correlation matrix will be of an immense help in the

study (personal discussion with Dr. Brunn).

 

2
9Blalock, Hubert M., Social Statistics, McGraw~Hill Book Company,

New York, 1960, p. 360.

CHAPTER III

DATA AND METHODOLOGY

A. Sources of Data

 

For the purpose of analyzing population change in the study area,
the following sources were used for the acquisition of data.

1. National Censuses: It has already been pointed out that the
availability of vital statistics on births, deaths and migration would have
been of great help to the author for the study. However, the absence of
such data was not completely a handicap since Ghana had population counts
starting from 1891. Census Reports are therefore available for 1891
(Southern Ghana alone), 1901, 1921, 1931, 1948 and 1960 (the country has
just finished with the 1970 Population Census and the data is being processed).

Though these reports are available their accuracy, especially the
counts carried out during the Colonial period, is in doubt. The early counts
were carried out by very few personnel and at a very little cost. With
the first three, the traditional method of counting people through village
chiefs was adopted and it was only in 1931 that the population was enumerated
individually.

The Chief Census Officer of the Ghana 1960 Census, remarked that the

1948 census had about 10 per cent under-estimation.1 This gives an idea of

 

1Gil, B. Z., West Africa, No. 2381, January, 1963, p. 68.

 

27

28

the extent of underestimation that might have taken place during the
earlier counts.

The year 1960 marks a milestone in Ghana demography for in that
year in March, a well organized population count was carried out. All
collected information was recorded separately for each individual of the
country. As a follow up, two months after the Census, a Post Enumeration
Sample Survey was carried out on about a 5% of the total population. The
objectives of the survey were (1) to measure "coverage and content errors"
and (2) to inquire into additional topics which could not be covered in
the main census such as household size and structure, internal and external
migration, religion, literacy, secondary occupation, marital status and
form of marriage, number of wives, fertility and mortality etc.

Data from this intensive and extensive 1960 Census was used in
conjunction with previous Census figures for the population change analysis.

2. National Archives: There is an office of The Ghana National
Archives in each regional capital therefore the Tamale and Accra National
Archives Offices were used for collecting relevant data. The author was
able to obtain material from some of the Departmental Reports and records
on the Colonial and independent periods; e.g., data on the peopling of the
area, the occurrences of the major diseases, population characteristics,
and general social and economic develOpment of the area.

3. Sample Survey: A Sample Survey was carried out by the author during
the months of January, March, April, May and July, 1969 in the study area.
He had the privilege of discussing his field research with Professor John
M. Hunter who was in Ghana from January to March, 1969.

Central South-Mamprusi has forty-six Enumeration Areas according to

the 1960 Census and these were used as the basic units for the research.

FIGURE 4: Central South-Mamprusi 1960 Census
Enumeration Areas

Enumeration Areas Numbered on the Map

 

1. Kpasenkpe 10. Wungu

2. Du ll. Logri No. 2

3. Arigu 12. Janga

4. Wulugu 13. Langbinsi

5. Zangu 14. Gambaga

6. Zangum 15. Kpakpirigbangu
7. Gbimsi 16. Nalerigu

8. Nayoku l7. Sakogu

9. Walewale l8. Namasim

l9. Nagbo

30

(Figure 4) A systematic random sample was taken with the result that
twenty-three enumeration areas were visited. In each of these, interviews
were carried out according to households.2 Considering the money
available, only twenty households in each.enumeration area were covered.
Thus, in all, four hundred and sixty households were interviewed, this
representing about 3 per cent of the total population.

The main objectives of the survey were (a) to obtain data which
could be used as supplementary material to the 1960 population census
Reports. (b) obtain data on population characteristics for the post-
censal period, that is 1960-69. To achieve these objectives questionnaires
were prepared with the help of Mr. S. K. Gaisie of the Demographic Research
Center, University of Ghana. (The questionnaires are attached as appendixes.)
With these, data was obtained on (a) fertility and mortality (b) in and out
migration with origins and destinations of both aspects noted (c) composi-
tion of household (d) occupational structure (e) marital status and (f), number

of wives and number of children.

B. Methods for Analysis of Problem

 

1. 1900-1948: The short comings of the early Census Reports have al-
ready been mentioned. Though.in 1931 and 1948 Census enumerators were
appointed to work in Enumeration districts no base maps were used. As
Monkhouse and Wilkinson have pointed out, "a properly conducted census must

3

be based on clearly defined enumeration areas." The lack of clearly

 

2A household refers to a number of people who obtain their food from
the same source. Since the extended family system encourages common bondage,
a household turned out to be a complete compound (house) of many individual
families.

3Monkhouse, F. J. and Wilkinson, H. R., Maps and Diagrams, University
Paper backs. London (1951), p. 260.

 

31

demarcated units makes it difficult to account for population change
according to the various sections of the study area. Thus for this period
the analysis is given at the local council aggregative level, that is,
considering Central South-Mamprusi together.

2. 1948—1960: The 1960 Census had two main aspects of operation
namely, the geographical preparations and the actual population count. The
enumeration areas, then serve as basic units for any population analysis
at the distributive level.

Though Central South-Mamprusi has forty-six enumeration areas, it
was thought helpful from the authors field research, to re-demarcate the
area into thirty three units with.boundaries coinciding in most cases with
the 1960 Census Enumeration Areas.4 In 1960, villages with a population
of one thousand and over were demarcated as enumeration areas, hence, I_
incorporated these into larger units to facilitate meaningful areal analysis.
(Figures 5 and 6)

The census material provide insufficient evidence to allow an
adequate distinction to be made between reproductive change (natural increase)
and gain from immigration. The Multiple Regression Analytic model was used
in order that the total variation of population change can be accounted for
in terms of other variables derived from the Census Reports of 1948 and
1960, mostly the latter, and the authors field research.

3. The Multiple Regression Analytic Method: It is a method which
consists of the measurement of the linear relationship or association be-
tween a dependent variable (Y) and two or more independent variables (X1, .

X2, X3, etc.) With this method it is possible to obtain multiple correlation

 

4Boundaries of Enumeration Areas coincide with Traditional cum
Administrative Boundaries.

32

.%o>uom vamfim m.uo:u:< cum momsou coma 0:050 "muuoom

m ousmfim

 

 

              
    

>.’h’* ‘—.“ ’h.‘." 'I'-‘

I’Cc.u o..P I

P

.a:! 5“ MW m. m “w

i

shoaso son

a. / .5010!
03:3;
/

    

>u>l3n 6:: 000- u..03h3¢ 02‘ nanZnU 60.— 5° «n.8(0230-

 

 

 

 

 

 

 

 

34
coefficients (R) which measure the estimated and observed value for Y
by the individual independent variables. In this way the research is
in a position to show the kind of relationship existing within the frame-
work of an established hypothesis or hypotheses i.e., the relationship
X

between Y and X X etc. respectively.

1’ 2’ 3

It is usually assumed that a correlation which is greater than
:_0.50 is significant and anything greater than :_0.70 is considered to be
high. However, this depends on the sample size. These limits are
arbitrary, and merely provide a convenient guide to making use of the
coefficients as descriptive statistics. In general terms coefficients
between -0.50 and -1.00 (perfect negative relationship) and between
+0.50 and +1.00 (perfect positive) are fairly significant; while of the
values lying between -0.50 and +0.50, little significant correlation is
to be expected. If a value of zero is obtained, this indicates that the
two sets of data fluctuate completely independently of each other and no
correlation exists at all.

In measuring up the relationship between the independent variables
and population change, the object is to explain why some enumeration areas
had higher population growth rates than others between 1948 and 1960. An
advantage of the regression model is that apart from indicating the type
of relationship existing between the dependent and independent variables it
can help the research know how much of the variance in the problem being
studied (in this case population change) through the use of the multiple
coefficient of determination (R2) has been explained, and how much remains
unexplained.

The multiple regression model is often expressed algebraically as:

Y = a-i-b1 Xl-i-b2 X2 + b3X3 + b4X4 etc.

35

where Y is the dependent variable (population change)

X1 " independent variable No. 1
X2 " " " NO. 2
X3 " " " No. 3
X4 " " " No. 4

a and b1, b2 etc. are constants.

4. Residuals from the Regression Model: The kinds of measures of

 

associations given above are only summary values: they apply to the
entire area under study and they cannot be used for mapping internal
variation of the degree of association. Their role is merely to describe
the degree of spatial correspondence among the areal variations of the
variables as measures of the validity of‘ the hypotheses for the area as

a whole.

The multiple regression model has an additional advantage of
providing residuals. A residual from regression for a particular observation
is the difference in magnitude between the observed value whose numerical
value is determined by factors included and omitted from the investigation,
and an estimated value, determined only by variables included in the
study.5 In this study a residual will be that part of the magnitude which
population change reaches in an enumeration area which is independent of
the areal association between population change and other independent
variables considered.

The map has been described as the fundamental instrument of geographic

research.6 It is possible to map residuals in an attempt to answer the

 

5Thomas, Edwin H., Maps of Residuals from Regression in Brian, J. L.
Berry and Duane F. Marble (eds). Spatial Analysis A Reader In Statisitcal
Geography. Prentice-Hall, New Jersey, 1968, p. 330.

 

6James, P. E. and C. F. Jones, American Geography: Inventoryyand
Prospect, Syracuse University Press, 1954, p. 9.

 

 

36
following questions:
1. Where are the large residuals found?

2. Where are the areas in which the regression equation
performs with reasonable accuracy?

3. Is there any recognizable pattern in the distribution
of the residuals that might suggest the effect of other
possibly unknown variables? (Hopefully, the locational
arrangement of the residual values will suggest other
variables which might be important in accounting for
the remaining variation in the dependent variable, i.e.,
the formulation of new hypotheses.)

Residuals have either a negative sign (in cases where the equation
over-estimates the value of Y) or a positive sign (in cases where the
equiation underestimates the value of Y). Thus a residual map will show
areas of over-estimation and areas of under—estimation and thereby depict

graphically the spatial variations in the performance of the equation in

accounting for differential population change.

CHAPTER IV

POPULATION GROWTH 1900-1948

A. Peopling of the Area

 

For a better understanding of the pattern in population change,
both at the aggregative and the distributive levels it has been considered
necessary to devote this section to giving an annount of the peopling of
the area before the turn of this century. The greatest event to take
place in the area was the peopling of it by groups of people from different
cultural reas. This factor of differential cultural background has through-
out the years affected many aspects of the human geography of Central South-
Mamprusi.
The area forms part of the Voltaic family and linquistically belongs
to the Mole-Dagbani association which covers a major part of Northern
Ghana and Upper Volta. It is known to have been settled for a very long
time perhaps before the neolithic period. The social organization of the
people was the same and leadership was vested in the divine representatives
of the Earth God (Tendana)*. Oral tradition does not help in telling us
how populous the region was but it can be inferred from the widespread notion
of the "tendana" that there was no vast areas unoccupied by man. Descendants
of these original people in South-Mamprusi are called "Dagbansabilisi."*
Upon these autochthonus people migrated small bands of people from
further north who were better armed, familiar with the idea of territorial

acquisition, and, in some cases, familiar with the rudiments of Mohammedianism.

37

38

The first group of 'invaders' to arrive were the Moshi-Dagomba
who claim to have migrated from Zanfara.1 The exact date of the arrival
of this group who later established the Moshi-Dagomba group of states
comprising Moshi, Mamprusi, Dagomba and Nanumba, has not yet been fixed, but
many suggestions have been made. It is most likely that they arrived in
the 15th century.2

According to oral tradition the leader of this group of migrants
from the North was Gbewa who after settling in Pusiga near Bawku encouraged
the movement of more men from further North. Due to succession disputes
after the death of Gbewa, Tohogu one of his sons, moved with.some of his
followers including Gruma to Mamprugu3 which later gave the group its
name. Gambaga was later on established as the traditional headquarters of
the Mamprusi. There is no evidence that the invaders received any serious
opposition from the original people apart from the Komkomba who opposed
them and that explains why there are no real Mamprusi in Komkomba.

Mercenaires and merchants have played a major role in the peopling
of the area. The oral tradition of the area has it that the first group

of mercenary-immigrants were some Mande-Dyulas who were invited by the

 

lThe Mamprusi, Moshi and Dagomba all claim to have come originally
from Zanfara, one of the Hausa Bakoi States. Tribal marks of the Zanfara
people who remained behind are similar to Mamprusi, Moshi and Dagomba
tribal marks. The Moshi-Dagomba group recognize the Zanfara people as
their "abokin wasa" (Hausa) meaning play mates.

2Nabila, John S., An Historical Geography of Central South-Mamprusi.
B.A. thesis submitted to the University of Ghana, Legon, March, 1968.

Mamprugu, now in ruins, was a village east of present Bunkprugu
i.e. North East of Nalerigu. The 1901 Census Report shows that it had
only 78 people. Mamprugu refers also to the Kingdom. (See the Glossary)

39

10th ruler of the Mamprusi Kingdom, Na Zibrim, in order to strengthen his
power. They were horsemen armed with spears and today their decendants
are to found in the royal towns of Nalerigu and Gambaga where they still
play the role of Spear bearers to the chiefs.

Na Atabia (1692-1752) went to Kong in the Ivory Coast by way of
Walewale, Daboya, Bole and Mango-Taro; and returned with a number of
mercenaries in 1751. These later on settled at a place they named Sansane-
Mango but some of them remained in Central South-Mamprusi and gave birth
to the Kanbonsi warriors of the chiefs. Their decendants are the chakosi
who speak a tongue understood by the Akans in the South.

The Tampolensi are Sisalas who arrived in Central South4Mamprusi
in the 1750's. Being mainly farmers they chose a fertile section north-
east of Walewale and established their own villages -- (1) Langbinsi, (2)
Singbini, (3) Ka-sape, (4) Boyene, (5) Zangua, (6) Nabilboka, (7) Tang-
bini, (8) Pinaba, and (9) Boku. (Figure 7)

The Mande merchants are known to have operated in the Gold Coast
for a long time, for the Portuguese met them at the coast in the 15th
Century. Their kinsmen, the spear bearers, invited them into Central South-
Mamprusi and they founded the following settlements: Nabari, Bulbia,
Guabuliga, Mushio and Zangu. A few of them settled in already existing
villages and built their own sections (Zongo), such as those in Gambaga,
Nalerigu and Wungu.

The Jihad of Uthman Dan Fodio coupled with the‘WarS 0f the Fulani
against the Hausa in Northern Nigeria in the early part of the 19th Century
forced many Hausa to migrate Southwards. Some of them settled in Central

South-Mamprusi and Binger saw them in Walewale and Gambaga.4

 

4Binger, L. G., Du Niger au Golf de Guinee, Vol. 2, Paris, 1892, p. 37.

40

.HH .a .mpma ..uau .mo ..m snow .maapmz

 

 

pmuu<

 

 

 

 

 

 

 

 

 

 

 

 

 

 

          

 

 

 

    

 

 

 

 

   

n whoafim
0.1.:
m. I M mI. . _
m A
1.11,“ . . =5. ‘50»
.1--- i . 77,131.03 Q .uclotohltfvoiauuo 3.2
III... . .32 :10 2» 0.. .. $8.8 £O~ c.3233 .todlré
Ill-l .Bocanzuln o...8€.8flilgngzun _
Il....::l.o..8.u= .. .. 009.133.1108...» _
>ux
.000 .90 u
. n
_
l
\
\ u I. _
\ . .. N \v... _
\ —\§§ —Il .\.
x .. 23%
. . a a Q
I I IIII . I r ILI . . IIIIAflo
l \ .1 \ a
\ . ’1 ‘ s a _
\ I I I ‘ OhgtlUW " ...q\\ Ohio
\ - .
\ X \ I a I 53.03»
\ s \\ $3. Jygfid 0..
. .
AI. II |||||||||| car H. w
III II, 0... a
at . .i If . \\ . a
’4 I .P‘ \\ o4 .......... . fl \“ ~\\ ’ ax .
I .o I
....... . . .X. _. x 3.5). 7.2225: - .,,., . .
2.23 1 .. .. \I .. . I. . 3.3.05
- \ \\ 1‘ — ll \\ \ 'I.
g u. s\\ \‘ III .1... II \ \ \ \ A a. h.
“ .mu;. 4 . ,, fl ,4.- _ m
. 1-..: . I.-. i a / 2.1.35: .
:93; R. I: as...» A]; \\ teal. w V .i/ .x/uuuu I ugh—mad. . I
’1 I \l A 90cc! all: ‘ \\. / x I.
I K I\\. \s I
a a 1 I\\ r .
.5.» o 333; a //z I” .
z ‘5 /.. I; K ..
IIIII III III I I no r OER
. ‘ ,
x X 30:20 N I I I ’ . can
’. 4; aJJSN O'OJQOW ”‘31.“? l I I 52" a
. I ~ ‘ I .
r r
.. x K“-.- ..\ . .I _:s oval-aha
I t. \ Ill I \ \‘
g .\ .. uzlaamFII 339dHVII 0'03 .
m .o...&- . .. V. .622... .
3x 0'30 _
I a
0.0 . ' fi/
. L
9.0 .I 9.0 - .90 I - I 00..

 

 

 

mFZmeotwm “.0 ZOEDmg—flo OZ< 20:30.3

Azsozm 03. 55; 20.3255

41

Through gradual acquisition of land, either by conquest or
settlement, mostly the latter, the Mamprusi Kingdom by the 19th Century
extended from present Chakosi in the East to about 1 40' Longitude West.
It had common boundaries with Sisala (Tumu) and Moshi in the North; Gonja
and Dagomba in the South; and W3 in the West.5

Political status accounted for the growth of the population of
certain settlements which became the seats of important chieftaincies.

A hierarchy of chieftaincies was developed with Nalerigu as the.seat of

the Paramount Chief, and Kpasenkpe, Wungu, Janga and Kurugu6 were divisional
headquarters. These became centers of attraction for social, political

and religious reasons for the people. (Table 1)

Before the turn of the century some Moshi men migrated from Upper
Volta in South-Mamprusi, most of whom were cloth dyers and weavers and
therefore settled in existing villages with markets such as Walewale,
Nalerigu, Gambaga, Langbinsi and Wungu. A few Moshi farmers are also
known to have migrated into the area at that time. Thus before 1900 the
population was made up of people from different places as shown in the

table below. (Table 2)

 

5Binger, op. cit. In 1888 he said that "Mamprugu controlled
Sesanimango and all the Grumshi up to the Red Volta," p. 38.

Kurugu was the divisional headquarters of the Talensi and
Bolatanga area. The 1931 Census shows a population of 1,427 for the
settlement. In 1933 Tongo and Bolgatanga were made separate divisional
capitals. As a result Kurugu lost its political function hence its popu-
lation reduced considerably. In 1948 for instance, the population was
only 70 and in 1960, 55; indeed by then it has completely declined.

42

TABLE 1

HIERARCHY OF TRADITIONAL AUTHORITY

Paramount Chief (Nayiri or Mamprugunaba)
Resident at Nalerigu

Divisional Chiefs

I

Wungu Kpasbnkpe Jahga Kurbgu
(No longer a di-

y vision headquarters
after 1933)

Sub-chiefs or Village Chiefs

I

 

I

Tendana and sectional leaders of Village

B. Population Growth 1900-1948

 

Between 1887 and 1901, the British, Germans and French engaged in
political manoeuvre in Northern Ghana. On 24th December 1896, the British
flag was hoisted in Gambaga and in 1901 the Northern Territories Protectorate
was created directly under the British Crown with Gambaga as headquarters --
Gamabaga lost this position to Tamale in 1907.7

With the establishment of 'Pax Britannica' the Slave trading ac-

tivities of Samori and Bakatu in Northern Ghana ceased. In like manner

 

7Metcalfe, G. E., Great Britain and Ghana, Documents of Ghana History
1806-1957, Thomas Nelson 1964, pp. 452—468 and 487-508.

43

TABLE 2

COMPOSITION OF POPULATION BY END OF
19th CENTURY - A RECAPITULATION

 

a,

Group Remarks

 

l. Autochthonus pe0p1e The indigenous people and main stock
of population today.

2. Those of Mande Origin These included Traouri, Diabakhate
and Kamara. Some of them formed the
chieftaincy group while others were
either merchants or spear bearers.

3. Hausa Immigrants Came to area after the Wars of Uthman
Dan Fodio.

4. Chakosi The Warrior group.

5. Tampolensi Settled in North-east of Walewale.

6. Moshi Immigrants Settled in area as traders and
farmers.

intervillage or inter-stage conflicts were stopped and all these gave an
impetus to free movement of people. Indeed there was a redistribution of

population in many parts of the North and adjoining Upper Volta.

1. Population Growth 1900—1921:

 

The methods used for counting people in the 1901, 1911 and 1921
censuses were very inaccurate, owing particularly to lack of qualified
personnel. As a result many villages were not enumerated and even in
villages where counting took place the response from the people was

poor. Even with the 1948 Census which was better organized than the previous

44

ones, the Chief Census Officer suggested that there was about 10% under-
estimation of that particular census.8 Even though one cannot rely wholly
on these figures they do serve as an approximation and as a guide to

what the trend was in population growth between 1900-1921. As the table

below shows, there was a steady increase in population.

TABLE 3

POPULATION GROWTH 1900-1921

 

 

Population Population
1901 1911 Change 1921 Change
Male 5,064 7,250 43.2% 10,200 40.7%
Female 3,920 6,609 68.6% 9,454 43.0%
Total 8,984 13,895 54.3% 19,654 41.8%

(After Census Reports)

 

Conditions of population growth during this period are not very
clear from the Census Reports. There was a population growth of 54.3 per
cent between 1901 and 1911 while there was a change of 41,8 per cent be—
tween 1911 and 1921. Though it is difficult to distinguish reproductive
change from net migration rate, it was the former which accounted for much
of the growth. This was during the early stages of the districts contact
with innovations introduced by Europeans and consequently birth rates were
quite high, thus conforming to the demographic transition theory. In view

of the population increases during the period, though this was a time health

 

8Gil, B. Z., 1963, op. cit., p. 68

45

conditions were very poor, it can be inferred that the fertility rate
might have been much higher than present rates. The family system to a
very large extent was based on the extended family system. Commenting on
the marriage in Northern Ghana as a whole, the 1911 Census Report has
this to say:
"All females after they arrive at puberty may be looked upon as
married, and males from an early age until they are old take upon
themselves as many wives as they can afford."9

Mortality rates were equally high in the society especially within
the young and old age group. Infant mortality was very high owing to
very poor health conditions and diseases like malaria, convulsions and
small pox took many lives from this young age group. Among the adult popu-
lation deaths were also common and it is known that the Influenza Epidemic
of 1918 raised the mortality rate especially within the age group 16 to 30
years. Cerebro Spinal Menengitis also caused many deaths especially among
females and the 1921 Census Report attributes the high sex ratio to the
deaths of many females as a result of this disease.10 The living conditions
at the time made many diseases endemic as well as highly epidemic in a few
places.

There was a general desire for many children especially males to
help on the farm, to tender livestock and to take part in other activities
associated with the subsistence economy. The people were mostly farmers
and in many places the wealth and prestige of a family was measured in terms
of the number of cattle, the extent of acreage under cultivation, and above

all, the size of the family. According to Caldwell, trepical African customs

 

9Gold Coast (Ghana) Census Report, 1911.

10Gold Coast (Ghana) Census Report, 1921.

 

46

and religions have always favored high fertility.11 The creation of
Gambaga as the first headquarters of the Northern Territories Protectorate
enhanced the position of Central South-Mamprusi. Even though this was
short-lived, an inertia in terms of attracting migrants into the area

was created.

When the British first took over the Northern Territories, they
found that whatever trade did exist was almost entirely in the hands of the
Hausa from Nigeria. The indigenous people were mostly farmers and not
interested in trade. More often than not these traders took cattle to
Southern Ghana in caravans and on the return trip brought back kola nuts
from the forest to the North. This helped markets in a few places, like
Walewale, Nabari and Gambaga, to increase in importance and consequently they
attracted immigrants. Even Walewale and Gambaga were important centers
for the trade traffic before the turn of the century. Commenting on the
importantce of Walewale in 1888 Binger said "Walewale :n; a sort of entrepdi
for Central Moshi, Sensanimango and North-Mamprusi."12 Thus the creation
of Gambaga as the headquarters along with the efforts of the British adminis—
tration to encourage trade resulted in the migration of some more Hausa,
Moshi and Mande traders into Central South-Mamprusi. The administrative
role of Gambaga also brought a few people from Southern Ghana into the area.
Table 4 shows the composition of the population of Gambaga by ethnic groups
in 1921. This table gives us an idea about the origin of most of the
migrants in South-Mamprusi. The density of population of South-Mamprusi

was 11 persons per square mile in 1921 (Figure 8). As this map shows there

 

llCaldwell, C., op. cit., 1967, pp. 39-40.

12Binger, L. G., op. cit., p. 52.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

II.

 

U
4
4
4

 

 

0-1
>1

 

 

 

D<>

 

 

”to:

I"'I

 

 

 

 

 

 

b 4* >04

 

t1 4

 

 

'r>

 

 

 

 

r<y<

 

 

 

 

b

 

>4}

V >

H>y4

 

I1

 

 

 

 

 

 

 

 

 

 

 

 

 

TY

 

oven :00
p. 38-5931“
100 -200 ’

FPP ATIN—Il.

4

 

as -so
pc_n so MILE

IO’ZS

 

SO - I00

 

 

 

O‘lO

 

 

 

 

After

Hilton, T. E.;

Figure 8

Ghana Population Atlas.

 

Thomas Nelson and Sons

Ltd. London. k960, p. 12.

ETHNIC COMPOSITION OF POPULATION IN GAMBAGA, 1921

48

TABLE 4

 

 

 

Ethnic Group Male Female Total
Mamprusi 375 438 813
Dagomba 7 15 22
Frafra 10 6 16
Grunshi 79 73 152
Builsa 6 8 14
Konkomba 5 29 34
Ashanti 9 0 9
Fante 12 3 15
Kusasi 3 4 7
Krobo 1 0 l
Kwitta 1 0 l
Dagarti 1 0 l
Bazaberimi 2 4 6
Hausa 64 86 150
Fulani 7 0 7
Lobi l 0 1
Kotokoli l 0 1
Yoruba 7 7 14
Moshis 32 28 60
Wangara 1 l 2
Total 617 702 l 319

3

(After 1921 Census Report)

 

was relatively a dense population in NorthrMamprusi. The fact of South-Mamprusi
being adjacent to a populous region was one of the causes for the migration of
some Grunshi, Frafra, Builsa, Kusasi and Busansi into SouthrMamprusi. Besides
this, there was the fact that South and NortheMamprusi formed one Administrative
Unit - the North Eastern Province — and the British officers encouraged people

from these areas to migrate into South—Mamprusi where land was abundant. Even

as late as 1950 John R. Raeburn estimated that in South-Mamprusi "only about

49

8 per cent of the total surface is cropped in any one year."13 Even
though this was the policy of the administrators only a few did move
because of the high attachment of farmers to their "fathers' land". By
1920 there were already Yoruba from Nigeria in the area and were definitely
engaged in commercial activities. In Gambaga for instance there were 14
Yoruba in 1921. (Table 4)

The above has been a consideration of conditions from 1901 to
1921 and the rest of this section will be devoted to analyzing conditions
of population growth from 1921 to 1948. It is hoped that the above analysis
will serve as the basis for the understanding of conditions in the rela-

tively contemporary period.

2. Population Change Between 1921 and 1948

 

This period is of paramount importance in the analysis of the
human geography of Northern Ghana as a whole and, indeed, in many other
parts of Ghana. Education was well established in the country and South-
Mamprusi boasted of a major Primary Boarding School in Gambaga (opened in
1912) and Primary Day Schools in Kpasenkpe, Janga and Wungu (all were
opened in 1947; the author was one of the pioneer students of the Kpasenkpe
School). Health conditions were improving with the building of dispensaries
in Gambaga, Walewale, Janga, Kpasenkpe and Wungu to cater for the schools
as well as the local population. There existed roads which served as

14
arteries for the important settlements, and there was even a proposal

 

3Raeburn, John R., Report on a Preliminary Economic Survey
of the Northern Territories of the Gold Coast, Colonial office, 1950,
p. 18.

14Nabila, John S., op. cit., pp. 41-42.

50

to build a railway line to the North.15 With the existence of motor
transport, movements became very common and this was an impetus for the
attraction of more migrants into sparsely populated South-Mamprusi. Of
more importance still during the period was the Spatial relationships

of South-Mamprusi, firstly with other parts of Northern Ghana and secondly
with Southern Ghana. Whereas it gained immigrants from the former, it
lost quite a few young men who migrated to the latter to seek employment.
0n the whole, living conditions were improving at quite a satisfactory
rate.16 All these developments during the period under consideration

had far reaching repercussions on population growth.

Between 1921 and 1931 there was a population change of 55.3 per
cent while a change of 34 per cent took place between 1931 and 1948. Both
Caldwell17 and Gaisie18 have indicated that birth rates have been very
high throughout the regions of Ghana for a very long time now. They have
suggested that fertility rates in Ghana have not changed very much from
what they were years ago. From this generalization, it is assumed for this
study that fertility rates in the various sub-divisions of Mamprusi by
1948 were almost the same as they were between 1911 and 1921. The inadequacy
of the Census Reports has made it difficult for any definite calculations

on fertility to be worked out. According to Caldwell, given reasonably

 

5Sessional Papers XXIV, Papers Relating to the Construction of a
Railway to the North from Kumasi, 1928-1929. The plan initiated by Governor
Guggisburg was scrapped by his successor because of the uncertainty that
the line could be Operated at a profit.

 

 

16Census Reports - 1931 and 1948 and Report on the Northern Territories
of the Gold Coast, 1937, Colonial Office, London (1937).

17Caldwell, John C., op. cit., 1967, p. 24.

l8Gaisie, S. K., op. cit., p. 39.

51

TABLE 5

POPULATION GROWTH 1921-1948

 

 

 

Per Cent Per Cent
Population Population
1921 1931 Change Change
Male 10,200 16,520 +62.0% 21,877 +32%
Female 9,454 14,003 48.0 19,104 36
Total 19,654 30,523 55.3 40,981 34

(After Census Reports)

 

good information on the age structure of a population and the distribution
of births by age of mother, it is possible to estimate the level of fer-
tility. Unfortunately, such information is not available on Central South-
Mamprusi for any estimations on fertility to be made.

By the 19305 the health conditions of the people had improved.
The building of dispensaries coupled with the existence of a relatively
satisfactory transport system made it possible for the rural areas to
receive medical care from Urban Centers. Though this was a factor that
helped in reducing the number of deaths in villages, the period between
1921 and 1939 was a very difficult time for the people because of the
occurrence of sleeping sickness.

Sleeping sickness, a protozoal disease, is caused by the genus
trypanosomiasis. African trypanosomiasis is a specific febrile infection
often characterized by weakness, wasting, and a protracted lethargy — a

condition which has given the disease its name, sleeping sickness. Unless

52

treated, the disease usually ends fatally.19

The prevalence of the disease depends on three major factors,
namely (1) the parasite or trypanosomes (2) the tsetse fly vector to
transport the trypanosomes and (3) a host (susceptible human or animal
population). Though the disease is known to be endemic in West Africa
it was never a serious disease until the 19205 and 308. It is known that
the first outbreak of sleeping sickness took place around Ouadougou in
Upper Volta but it was restricted to this area until the Colonial period
when it spread to other places.20 Lack of movement among the people during
the pre-colonial period was a major factor in curtailing the disease.
Studies have shown that the appearance of some epidemics followed directly
the need among the people for trade, employment or the occupation of new
land. By the 1930s Southern Ghana had already become a receiving area of
labor migrants from Northern Ghana, Upper Volta and other neighboring
countries. The influx of migrants from the North to the South did not
only introduce the disease in Ghana but also raised the incidence rate
of sleeping sickness. Checks at the Yeji ferry revealed that most of the
migrants to Southern Ghana were infected with sleeping sickness.21

As Figure 9 shows, SouthéMamprusi was one of the areas which suffer—
ed from sleeping sickness in Ghana. The development of the focus was

largely due to the permanent settlement of Moshi in SoutheMamprusi and of

 

9Strong, Richard, Diagnosis, Prevention and Treatment of Tropical
Diseases, New York, 1945, p. 164.

 

20Morris, R. S., "The Ecology of Epidemic Sleeping Sickness, The
Significance of Location." Part I and II Bulletin of Entomological Research,
Vol. 43, 1951.

 

21Scott, David, "The Epidemiology of Sleeping Sickness in Ashanti
(in three parts) Journal of Tropical Medicine and Hygiene. Vol. 60, 1957.

 

 

 

53

INCIDENCE OF SLEEPING SICKNESS IN GHANA 1919-1939

—_— 4-

 

 

.—.

 

 

 

   
   

 

 

 

 

 

 

 

 

 

3° 1°, WW 1%;
H°‘ I l '1... I N
.. E R oBOloton }
........ ' . . ..:.:1:: __4
Ni
I I N ——|0°
’J
E ” 5"
”‘9. § __
\u
5 R a:
i - t
z _90
9o... Damongo
\ I, __
\\ .. Z
\..
80,... 0 ._...8°
8 R o N A H A F o \ 0
\
rs I’4‘-’"_"\\ \\ In _
Sunyani o ‘\“'l 63459, '~\ 4 (I
'10—— Ir’ASHATl/ /( __70
‘b
\‘ 9‘ r’ l , .—
\ / oKumosi I ( 3H0
19 \\ I l y
_ \ f
Jl REGION/’EASTERN O
I s 7‘ I x
s ‘ A >
——1 be
.__1
-—:5°
‘1’ 3°“
'6
Foci of sleepifl Sickness L

 

 

Figure 9

After David Scott; "The Epidemiology of Sleeping Sickness in Ashanti"

(in three parts) Journal of Tropical Medicine and Hygiene, Vol. 60,
1957.

54

course,a favorable environment for the development of the disease; the
existence of the Tsetse fly vector and a susceptible human population.
what was absent was the parasite, trypanosome which the Moshi immigrants
supplied. Since the 16th century the Moshi, Mamprusi, Dagomba and Namumba
regarded themselves as sister kingdoms. With the development of free
movement of people.coup1ed with the existence of abundant land in South-
Mamprusi,tfiuaMoshi felt well accommodated to settle in the area. At times
they established their own villages, for example, Sayo, Nagbo, Banawa,
Manga and Guabuliga. Indeed, the authors field work has revealed that
there are Moshi immigrants in almost all the settlements in Central-South
Mamprusi. Records in Gambaga and Tamale reveal that the disease was
unknown until many Moshi made the area their home. Though it is difficult
to tell the exact number of people who died from this disease, Table 6
clearly shows South-Mamprusi with the highest incidence rate in Ghana as

a whole between 1932 and 1939. Through the efforts of the Colonial Adminis-
tration the disease was eradicated and today the Rural Health Department,
with a head office in Gambaga, has the duty of making sure that the people
are free from the disease.

Apart from the incidence of Sleeping Sickness during this period,
the area faced periods of famine owing to the invasions of locusts. The
first invasion took place in 1929—30 and it has been recorded that 545 farms
were destroyed by the locusts.22 There was a second invasion of the locusts
in 1933/34 and a final one in 1939. During the author's field work, the

people gave vivid accounts of the hard times they experienced due to these

 

2
2District Dailnyiaries, Gambaga Administrative Office, Gambaga.

INCIDENCE 0F SLEEPING SICKNESS
(Trypanosomiasis) in Ghana 1932-1939

55
TABLE 6

 

 

All Kumasi

Year Ghana Hospital Sunyani Mamprusi /Tumu
1932/33 685 93 * * *

1933/34 1.179 363 73 250 1
1934/35 1,973 252 96 1,012 150
1935/36 3,885 249 95 1,683 1,161
1936/37 4,820 411 119 2,323 944
1937/38 5,594 306 155 2,148 1,242
1938/39 5,611 258 36 1,415 1,107
1939/40 6,826 * * * *

 

South Mamprusi and Lawra/Tumu were endemic areas with incidence
rates of 6% and 3.5% respectively.

*Figures unrecorded.
After David Scott, "The Epidemiology of Sleeping Sickness in

Ashanti," (in three parts) Journal Tropical Medicine and Hygiene, Vol.
60, 1957.

invasions and, of course, the sleeping sickness epidemic. With the foregoing,
it is quite evident that death rates between 1921 and 1948 might have been
higher than they were in the previous years.

After 1921, immigration mainly accounted for the increase in popu-
lation in northeastern and northwestern areas of Northern Ghana. These

immigrants who were seeking land for cultivation or were desirous of

escaping the stringent political and administrative (tax) systems in their

56

respective countries were mainly from Upper Volta, northern Togo,
Mali and the Ivory Coast. About 50 per cent of them settled in north
Mamprusi, therefore, by the 19305 there was already a problem of over—
p0pu1ation.23 There was a steady influx of immigrants from North-Mamprusi
into South-Mamprusi. These immigrants were mainly agricultural settlers.
In addition to the immigrants from the north, there were some Fulani
who are noted for rearing cattle very well in some West African countries,
e.g., Nigeria, Senegal and Guinea. The British administration, at the time
also in charge of Nigerian affairs, encouraged the Fulani to migrate to
Northern Ghana to help farmers keep their cattle. This, they hoped, would
produce better cattle as well as allow Northern Ghanaian farmers to con—
centrate on food production.

In 1945, the Gold Coast (Ghana) Department of Agriculture held
a survey in South—Mamprusi and as a corollary to the survey, four villages
were visited where immigrants were interviewed. The results of the
interviews, as shown in Table 7, indicate quite clearly the role which
inter-regional immigration from North-Mamprusi played in the population
growth of the area.

Fishing in rivers, streams and ponds has been well established
in the area. However, the Mamprusi usually engaged in fishing during only
the dry season when the water was low, especially between the months of
September and May. Kpasenkpe, Arigu, Janga, Mushio, Bukpere and Nasia have
long been associated with fishing, but in none of these areas, was it a
serious profession. In 1929-30 some Battor immigrant fishermen from Tongu

villages on the lower Volta, where they faced problems of over-population

 

23Dickson, K. B., op. cit., p. 280.

57
TABLE 7

IMMIGRANTS FROM NORTH—MAMPRUSI
AND UPPER VOLTA, 1945

 

 

Average

No. Yrs.
Village Frafra Moshi Grunshi Busanga Total Settled
Langbinsi 9 17 20 2 48 12
Sameni 12 l 7 - 20 15
Nayoko 4 9 11 6 30 8
Gbimsi 13 26 25 - 64 10

Total 38 53 63 8 162 —

 

*The number of years settled varied from 2 to 25 but for the most
part corresponded nearly to the average figures given above. Grunshi,
here, refers to Builsa and Kasena. Moshi and Busanga were mostly from
Upper Volta.

After Smith, R., Agriculture in South Mamprusi, Agriculture
Department (1945), p. 9.

and increasing proverty in the home area, arrived. They settled in almost
all the fishing villages mentioned above and in one instance they found
their own settlement -- Mafanga. Since then, the fishermen continue to
immigrate into South-Mamprusi and they carry out fishing on commercial
basis along the White Volta.

The increasing importance of commercial activities in the district
brought more Yoruba, Hausa and quite a few traders from Southern Ghana.
Some Dagomba traders and farmers also immigrated to settle in the district.
Another source of immigrants was the eastern section of the district where

live the Bimoba and Komkomba. A few of these immigrated to some of the settle—

58

TABLE 8

NUMBER OF PEOPLE ENUMERATED FROM NORTHERN TERRITORIES
IN THE COLONY AND ASHANTI, 1948

 

Kusasi 2 Sisala 4
Nankani 1 Dagarti ll
Mamprusi 2 Lobi 7
Bimoba 0 Wala 14
Kassena 2 Gonja 3
Builsa 3 Dagomba 7

Konkomba 2

(After 1948 Census Report)

 

ments in eastern Mamprusi for example Gambage, Nalerigu, Sakogu and Langbinsi.
Data from the Census Reports for the major ethnic divisions from
Northern Ghana but enumerated in Southern Ghana in 1948, gives some in-
dication of the long distance mobility of the various groups.
Though it is most likely that there was some under-estimation
of the actual number of people, the figures do suggest that there was a
small movement of labor to the South from Mamprusi. In like manner out-
migration from the area to other parts of the North was very insignificant.
The only out-movement of any sort to other parts of the North was through

a few marriages of Mamprusi women outside the district.

CHAPTER V

POPULATION CHANGE - 1948-1969

A. Population Change — 1948-1960

 

The 1948 and 1960 census figures are more reliable than those
of earlier periods. In 1960 the total population was 61,372 giving an
overall population increase of about 54 per cent, i.e. between 1948 and
1960. This gives an intercensal growth rate of about 3 per cent per year.
Table 9 gives the population for the four traditional-administrative
divisions. In it, the per cent change in population and population densi-
ties for each division has been given as a basis for comparative analysis
at the divisional aggregate level.

Birth rates continued to be the same while the death rate was
reduced very considerably with the opening of a modern hospital at
Nalerigu in 1957 by the American Baptist Mission. Immigrants continued
to settle in Central South-Mamprusi so that by 1960 almost all the major
ethnic groups of Ghana were represented (Table 10). As Figure 10 shows,
in 1960, South—Mamprusi was one of the intermediate zones lying between
the sparsely populated middle belt and densely populated north—east and
north-west.

The analysis of population change between 1948 and 1960 is based
on the thirty-three population unit areas (Enumeration Areas) which were
derived from the Ghana 1960 Census and the author's field research.

(Figure 6)

59

6O

DENSITY OF POPULATION-NORTHERN GHANA (1960)

 

  

 

 

 

 

I A. I
2. 11. E E . 00
mo --- -;-—-—_---~_....._._...J:‘ - - ...:: ..... fiE 11“
— — _ ................ \.
__ _ __ ‘.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P103; P . . . — — — —(. 101‘
K. J'. J; 7 _ _ _ _!
‘ “Y j__ _ _¢.J

(' ' ' ' ' V 1 ._ _ .._'.‘.’.’.‘_1_
. . . . 1
‘. . — "" _ '— ‘i
./. . . . , _ _ _ _‘
.’. ,' . . -— — —l
"'9. k} . . . . -— -— -— - '\ 9°-
'\. . . . — — _ _ _ \

 

 

 

 

 

 

 

 

1 \\ 5‘. ‘— --—— —— —— __ J~_’ \
1 \x. 1"; \‘_ 7-- s — ) ~r‘ I"
j \“v’ ‘ _’/ \\_‘ Q!" ’
f' |‘_( 5.:
' K
s8° k...) g ‘ 83-
L'. ' ' 10 0 10 50
0 1O 25 50 75 100 200 300 MILE S

 

f-‘rom1960 Population Census of Ghana.

Figure 10

61

Amuuoamm msmcoo Houw<v .ummz onu aw mum mfioamfi>aa «excommmx can mmoww .swcsz osu mo muuwm .muswfim
mufimomw mfinu cmuomwwm manmaum>aw w>m£ Aomm.av =wc=3 can Amwc.qv mam3oam3 Mom mouswfim aofiumasmoms

 

 

 

 

 

 

om mm Am><vw.sm Nam.as soo.s~ moq.mm Hwa.oq soa.mfl smm.H~ Nmo.H assoc
m as mm moo.s mmH.H omm.~ cas.~ 00H.H oao.a qsa «mums
Nm no as mmw.m moa.m omm.s sma.m osm.~ sao.m owa mascmmmas
sq Hm won «NH.mH «No.5 ooH.w sma.¢a oom.s «mm.m saw swans
New Nms H.mm 0mm.mm omo.oH oom.~a oma.- ss~.oH oHo.HH NmH.H awfiHmHmz
m z
owomnu A.fia .omv mufimcoo Hmuoa m z Hmuoe m z moHHa .om aowmfi>finnn=m
cowumasaom oosfl msafl mmu<

 

coma 92¢ w¢ma mZOHmH>HQImDm rm ZOHH<ADmOm

m mAm<H

62

mnoo< .qomfl ..m uuoaom Hafioomm .msmaou soquHamom coma umum<

 

oHN
omN
OHN.sH
mso.H
on
omo.H
ooo.a

qu.m

masouw ofianuo an vofimwmmmao uoz own
msmowuu< nonuo coo
mazuo 0mm
fiamasm HH
Haemaom OHN
ovum: no
anwcsuu own
Aamsuaamz mcwwsaoasv acmpwmanmaoz con.~m

moan:

mnsuow

omsu

mmmflowoa Hmuuomo
63m

onwomo<lmu

awx<

wmsumsmz

 

masouo owanum may
Hmsmmz<zwmaaom A<MHzmu zH onH<A=mom mo onHHmomzoo

OH mqm<H

63

A.muuomom msmooc coma cam wqmfi mcmscc
H unawam

 

 

On. 0. Oh

5.9.08

 

afloao
Onoosuc.
u02<zu

 

«l9. 0

uO.—¢¢
I. D [P L
9 ‘ q j
0—
COUCBUCO

 

Goo—Invo—
ugv,.(.¢u .5nv.h.< i.e.-nun

 

 

64

Figure 11 gives the population change in the different areas
between 1948 and 1960. As Appendix B shows, of the 33 areas, 12 had high
growth rates, 8 had moderate growth rates, 6 had low growth rates while 7
had a decrease in population change. An attempt is made to account for the
differential growth rates in these areas in the rest of this section.

1. Results of the Multiple Regression Analysis

(1) Preliminary Considerations: The variables for the analysis have
already been given in Chapter 2. For the purpose of the analysis, the
author designated ranking values of 1,2,3, and 4 for village chief, sub-
divisional chief, divisional chief and paramount chief respectively, i.e.
values for traditional authority. A settlement with a population of
1,000 was considered as an urban center and was awarded 1 point. For
example, Walewale with 4,493 people, got 4.5 points, which was the highest
value for these urban centers. For the other twelve variables the author
utilized the Census Reports to obtain values for them.

The data used for the regression model was in nominal, ordinal
and ratio scales. Apart from this, there were great variations in the
values for the various variables considered in the study.1 It was there-
fore thought necessary to carry out a log transformation. Blalock has
pointed out that where the independent variables take on a wide range of
values such that a certain value is reached further increases produce less
and less effect on the dependent variable. This could even result in a

non—linear relationship.2 The logging of the figures will have the effect

 

1
There were a number of extreme cases where some Enumeration Areas
scored zeros for per cent born in another region and foreign born.

2Blalock, H. M., op. cit., (1960), pp. 312-313.

(i)

65

of bunching together the extremely large scores and lessening the bend
effect of the large scores. One major prerequisite of a correlation

and regression model is that there should be some measure of a linear re-
lationship between the variates, i.e. a normal distribution. Indeed, a
preliminary treatment of the raw data without any transformations resulted
only in an R2 of 15% for all variables. A transformation of logarithms
(X1=log. 10X) were thus derived for the analysis, and this transformation

was performed on all variables.

Factors Accounting for Population Growth

 

An examination of the simple correlation matrix (Appendix 1) and
Table 11 reveals that only three of the independent variables have no
significant correlations with population change namely, per cent born in
another region, sex ratio and dependency ratio. The table gives correla-
tions between the independent variables and population change, and their
levels of significance, plus the standard partial regression coefficients
which are the 'b' values. Each 'b' value gives the rate of change in the
dependent variable for a unit change in the particular independent variable
while remaining independent variables are held statistically constant.3
Thomas, for instance, has said that these figures do indicate the individual
contributions of the variables towards explaining the variation in the
dependent variable.4

From Table 11, in order of importance,the regression coefficients

of traditional authority, urbanization, education,another locality, and

 

3King, Leslie J., Statistical Analysis in Geography. PrenticeéHall,

Inc., New Jersey (1969), p: 140.7

4Thomas, E. N., op. cit., 1960, pp. 158-170.

66

.ucmoamwawam ooz«

 

 

 

we. Hc. w~.c ow. coaumosom .MH
so. « No.H as. oaumm soaouamama .NH
mm. Hcc. «H.mN no. oOHuMNfiamnua .HH
NH. Ho. m¢.HH Hm.l .omnus scum monoumflo .cH
oc. Hcc. ~¢.NH co.I momou aoum oosmumwn .m
on. Hcc. me.mo No. mufinonu=< Hmooauwomue .c
ma. « so.H mu. owumm xmm .s
cm. Ho. q¢.HH Nm. anon awaouom N .o
«0.: « Nc.H cH.I coawou nonuonm I :uom N .m
on. mo. ms.m mm. suwamoofl “masons I anon N .s
ca. mc. cm.m NM.I huwamooa manu I anon N .m
mo. no. mm.s on. owumm suaawuumm .N
am. Hoe. ¢~.mm Nu. suHmcoa .H
muooaofimwooo oocmofiwwcwfim oSHm> owcmsu
cowmmouwom mo Ho>oq moamoamwcwfim .mom on“: oHomem>
Hmfiuumm oumocmum nowumamuuoc
HO. MH " HQPGH HOG. : : z = .. : : AMV
Gm. h u H0>QH HO. : : : : : .. ANV
mc.¢ u Ho>oH no. on» on woomofiwwawfim wow osam> .m Hmowufiuo AHV

.mmou< coaumuoascm mm ommo was» as .moowum>uomno mo Hones: onu wH z can aooooum mo ooumoo on» ma m oumnz

u u H

Assn .a ..uso .ao ..z.m scoumsm Imuu<v A .28 «I H -
N I N I z

.Hm umoH Mom mH55hom

 

mHzmHonmmco onmmmmumm A<HHm<m Qm<nz<Hm mmH Anv
924 mchH<AmmmcU mamZHm may no moz<UHmHchm Mom Hmma muz<Hm<> mo mHqu¢z<

HA mamde

67
TABLE 12

SUMMARY TABLE FOR THE STEPWISE
REGRESSION MODEL (LSADD AND LSDEL)

 

 

Posi— 2 Increase or
tion Variable R R Contribution
1 Traditional Authority .8207* .6736 .6736
2 Urbanization .8628* .7445 .0709
3 Education .9055* .8199 .0754
4 another Locality .9184* .8436 .0237
5 Density .9261* .8577 .0141
6 Foreign Born .9306* .8661 .0084
7 Sex Ratio .9340 .8723 .0062
8 Distance from Roads .9361 .8764 .0041
9 Born in this locality .9389 .8816 .0052
10 Dependency Ratio .9397 .8830 .0014
11 Distance — Urban .9400 .8836 .0006
12 Fertility Ratio .9403 .8841 .0005
13 Born Another Region .9403 .8842 .0001

 

*Significant at the .05 level.

foreign born contributed more toward explaining the total variation than the
other variables. As an aid to understanding the individual contributions of
the variables both step-wise deletion and addition ("Least Squares Deltion"
and "Least Squares Addition") programmes were carried out with the logrithmi-
cally transformed data. The results have been combined and shown in Table 12

in order of importance. In Chapter II, it was envisaged that the simple

68
correlation matrix should help in bringing out meaningful demographic
patterns in South-Mamprusi. Indeed, there are more higher inter-correlations
than those between the independent variables and the dependent variable,
and these will be utilized for the analysis. With the foregoing, a con-
sideration of the independent variables individually will be given in order
of their magnitude in an attempt to account for population change as well

as test the hypotheses formulated at the beginning of the study.

(a) Traditional Authority; This has a simple correlation coefficient

 

(r) of .82 with population change and is significant at the .001 level. It
is the most important variable related to population change. It has an

R of .8207 and R2 .6736. Thus, it "explains: by itself 67% of population
change, and points to the importance of cultural considerations on popu-
lation growth. The hierarchy of traditional authority has already been

noted and the role of chieftaincies in attacting local and outside immigrants
has also been noted earlier. It was hypothesized that there should be a
direct relationship between traditional authority and population change, and
the results support this assumption.

In considering fertility levels in Ghana as a whole, Caldwell has
pointed out that Ashanti has the highest fertility rates because of the
traditional system, apparently derived from the former Ashanti Kingdom.5
Both Gaisie and Caldwell attribute the low levels of fertility among the

ethnic groups in Northern Ghana to sex imbalance and seasonal and long term

migrations which tend to separate couples for short and/or long periods.6

 

5Caldwell, op. cit., January 1967, p. 222.

61bid., p. 222;

Gaisie, S. K., op. cit., 1969, p. 47.

69

(Table 13) Though this is the case at national level, there is evidence
to show that the Mamprusi engage in very little long-distance migration.
(Table 14) Though there are no sociological, ethnographic and demographic
studies on the major Ghanaian ethnic groups, there are some basic considera-
tions which the author found during his research within the traditional
system to be of relevance to population change.

The Mamprusi just as the Dagomba, Nanumba, and Gonja and Wala
in Northern Ghana has a centralized political system with rotational
succession to political offices. In contrast, the other ethnic groups
in the north have long standing of acephalous societies. It has been pointed
out in Chapter IV that the "royals" are descendants of immigrant conquerors,
while commoners are the autochthones. In Mhmprusi, the royals constitute
about one-third of the total population. Socially and economically, the
people are differentiated into groups of royals, warriors, Muslims,
musicians, craftsmen, and commoners. Fortes noted that these centralized
societies have "centralized authority, administrative machinery, and
judicial institutions, and in which cleavages in wealth, privileges and
status correspond to the distribution of power and authority.7

Status in the traditional society implies, more often than not, the
acquisition of many wives and children. On the whole, polygamy is wide-
spread in South-Mamprisi, and children are regarded as an economic asset.
The perpetuation of the line of descent is of paramount importance to any

family. Among the Talensi, Fortes said "there is something wrong by custom

 

7Fortes, M. and Evans-Pritchard, E. E., African Political Systems,
London, Oxford University Press, 1940, p. 7.

70

TABLE 13

GHANA, TOTAL FERTILITY BY ETHNIC GROUP AND URBAN/RURAL

 

 

 

Ethnic Group Total Fertility** Urban Rural
1. Akan 6.6 5.5 7.1
2. Asante and Ahafo 7.3 5.5 7.7
3. Fante 6.5 5.3 7.3
4. Nzema 5.2 5.2 5.2
5. Akyem 5.9 6.2 5.7
6. Akauapem 5.2 3.7 5.8
7. Kwahu 7.3 7.2 7.5
8. wasa 5.3 ---* 5.5
9. Boron (including Buda) 7.2 5.8 7.4

10. Ga-Adangbe 5.8 5.5 7.4

11. Ga 5.6 5.7 4.7

12. Adangbe 5.8 4.2 6.1

13. Guan 6.3 5.4 6.9

14. Ewe 6.6 5. 6.8

15. Gurma 5.6 ---* 5.4

16. Konkomba 4.4 ---* 4.7

17. Lobi 6.2 ---* 6.2

18. Grusi 5.0 ---* 5.0

19. Central Togolese 6.4 ---* 5.4

20. Mole-Dagbani 5.3 5.2 5.4

21. Dagomba 4.8 5.0 5.0

22. Mamprusi 6.3 ---* 6.2

23. Dagarti 6.6 7:6 7:2

24. Builsa 5.1 ---* 5.1

25. Frafra 4.0 3.0 4.1

26. Kusasi 5.2 -—-* 4.8

27. Moshi 5.9 6.5 4.9

28. Yoruba 6.1 4.9 6.4

 

Source: Gaisie, S. K., Dynamics of Population Growth in Ghana,
University of Ghana, Legon. 1969, pp. 39 and 43.

 

*Insufficient information for the computation of total fertility
rates.

**Total fertility is the estimate of the number of children a
cohort of 1,000 women would bear if they all went through
their reproductive years.

71

TABLE 14

SHORT AND LONG RANGE MOBILITY BY ETHNIC GROUP
AND SEX, NORTHERN GHANA: COMPARATIVE ANALYSIS

 

Per cent of people Enumerated

 

 

 

 

 

 

In Another In Another
Locality, Region of
Ethnic Group Same Region Ghana Abroad
M F M F M F
Group A: Centralized
States
Mamprusi* 12.9 24.7 8.0 3.0 0.3 0.1
Gonja 17.8 22.0 7.3 8.3 0.1 0.1
Dagomba & Nanumba 20.8 31.7 11.5 9.3 0.1 0.1
Wala 12.3 26.5 14.1 10.6 1.0 0.8
Groupr: Acephalous
Igpcieties
Komkomba 26.0 31.7 12.8 12.4 4.8 4.9
Dagati 11.1 37.3 20.4 8.6 5.5 5.0
Builsa 4.2 29.5 15.8 7.3 0.2 0.1
Grunshi 3.6 43.2 0.5 0.1 0.1 2.2
Frafra 9.0 38.5 26.3 10.8 0.8 2.0
Talensi 4.8 37.3 1.7 0.6 0.1 0.1
Kusasi 7.2 16.0 15.6 6.8 1.3 1.1
Sisala 8.3 47.6 15.0 7.5 9.0 1.6
Kasena 5.6 47.2 2.2 0.9 1.6 7.5
Moshi and Vagala 11.4 20.4 37.7 23.7 19.3 14.8
Lobi 4.7 8.5 2.8 2.1 44.4 40.4

*The figures include the Eastern and Western South-Mamprusi.
(Source: 1960 Population Census of Ghana Special Report E., 1964, Accra.
Table 4, pp. 25-26.

72

with men and women who never marry, and they are few."8 This is very
true of the Mamprusi, especially among the royals who are made up of
chiefs and their sons, elders, courtiers, and others associated with
authority. These are noted for the possession of very large families.

Moslem populations usually depict higher fertility rates than
other major religions. Kirk has attributed this to characteristics of
Muslim institutions which favor high fertility.9 At the time of their
migration into Northern Ghana from further north, the royals already knew
some rudiments of Mohammedanism. Since about the 16th century, the
centers of chieftaincies and commercial centers have often attracted Muslim
immigrants. Since Muslims are an integral part of a Chief's court in
Mamprusi, the traditional authority capitals must show a high level of
fertility--higher, say, than the villages with or without Muslims.

An examination of Table 14 reveals that there are some differences
in the mobility of the people between the groups based on traditional
authority--centralized and acephalous. The movement of males intra—
regionally is significantly greater than that of males in the acephalous
societies; the average per cent of males enumerated in a locality other than
their natal one is double that of group B societies. The men in the
centralized societies migrate more locally than they do outside the region.
In contrast, men in group B move very little locally and when they move

they do so outside the region (inter-regional migration).

 

8Fortes, M., The Web of Kinship Among the Tallensi, London, 1949,
p. 83.

9Kirk, D., "Factors Affecting Moslem Natality," in Bernard

Berelson (ed) Family Planning and ngulation Programs, Chicago, 1966, pp.
561 ff.

73
Regarding female local mobility, census figures show that in
every locality more women than men are living in localities of the region
other than their birth place. On the whole, in the acepalous societies,
lineage and clan exogamy and virilocal marriage are the rule. In contrast,
in the centralized states there are no prohibitions about wives to be
chosen from distance places. The states are characterized by a complexity
of many ethnic groups and do not therefore cater for exagamous socieities.
From the above, the role of traditional authority in controlling
the movement of people is quite obvious. The hierarchy of chieftaincies
do generally generate intra-local migration. The successor to the paramountcy
is usually chosen from among the chiefs of the "skin-villages" in Mamprusi.
At the lower level, in the traditional system, smaller chieftaincies are
stepping-stones to higher political office. Others, like the divisional
chiefs of Kpasenkpe, Wungu, and Janga are termini at which.the chief reaches
the limit of his ascent in traditional authority.10
Customarily sons cannot reach higher positions in traditional
authority than fathers. In pursuit of a higher chieftaincy, an ambitious
chief's son goes from office to office of increasing rank until he reaches
the limit of his line. The ascent of this ladder of political office often
involves frequent movement from village to village over considerable dis-
tances. Such a movement does not affect the chief and his immediate family
alone, but also many others, including patri- and matri-kin, personal members

of the court, clients, specialists like musicians or craft-specialists.

Nalerigu, the seat of the paramountcy, is noted for having people from

 

10The Mamprusi have a saying that "The Prince has no home village,"

indicating the high incidence of mobility among the royals.

74

more than one-third of the villages in the east. There is a case of
Na Wafu of the Mamprusi State who is said to have moved six times, i.e.
to six different villages, before becoming the paramount chief in
Nalerigu.

The traditional political system serves as a check on the migration
of people to other regions. With such a system, many men are bound to be
involved in the whole set-up. Besides, there is the fact that the system
calls for very close kinship ties, such that only a few people can really
get away from home for long distances, especially with the purpose of
settling there permanently. Perhaps this partially explains why the

Mamprusi engage very little in long-distance migration.

(b) Urbanization: At the national level, urbanization is unimportant in

 

Central South-Mamprusi. However, within the area itself some settlements
like Walewale, Langbinsi, Gambaga, Nalerigu, Wungu, and Sakogu do have quite
sizeable populations. It was hypothesized that urban centers are areas

of rapid population growth. Urbanization has a simple correlation value

of r=.67 with population change thus showing that it is significant at the
.001 level for the study. It contributes about 7% of the total variation

in the stepdwise regression model.

A few studies in Ghana and elsewhere have revealed that there are
fertility differentials between the rural and urban populations. Fertility
is usually fouhd to be higher among rural residents than among the residents
of towns. The rural fertility exceeds the urban fertility by 15% in Ghana.11

From TablejL3,the role of urban centers in reducing fertility among Ghanaian

societies is clearly shown. However, the table also reveals the low level

 

llGaisie, S. K., op. cit., 1969, p. 29.

75

of the degree of urbanization in Northern Ghana. Caldwell found that
Northern Ghana, Volta Region, and Brong Ahafo are the least urbanized
regions in Ghana.12 A further consideration of Table 13 still indicates
the absence of this phenomenon in many Local Councils, SouthéMamprusi
being one of them. What is happening is that Urbanization is still
developing in the few villages mentioned above. As a result, urbanization
is rather playing a reverse role here, i.e. instead of reducing fertility,
it is raising it. A few of the urban centers are also traditional authority
capitals eg. Nalerigu, Wungu, and Gambaga. There is a simple correlation
value of r=.55 between Urbanization and traditional authority, and what
has been said about the latter is very true of the former in terms of the
roles they play in increasing the population.

The few urban centers have either medical facilities (a hospital
in Nalerigu, a health center in Walewale and dispensaries in the others)
or easy access to hospitals outside Mamprusi. This has been a contributory
factor towards the lowering of mortality rates, especially infant mortality.
On the whole, living conditions in the urban centers are far better than

what it is in rural areas and they continue to serve as attractive centers

for the local populace or immigrans from other regions.

(c) Education: It has been noted that a survey in Ghana revealed that,of
the males and females who had never been to school, 27% and 16% respectively
migrated at some time to an urban area; but 67% and 61% of those who had
high school or university education had done so. This being the case,

education should be a significant factor for population growth. Most of the

 

12Caldwell, "Population Change," op. cit., 1967, p. 132.

76

immigrants from other regions and are engaged in administration, commerce,
education (teaching in schools and a college) working in the hospital and

the like have all received some sort of education. Urban dwellers have a
higher proportion of literates than what exists in rural areas. The variable,
number of people with present or past education, has a correlation of .46

(r) with population change and this is significant at the .01 level.

A few of the urban centers and traditional authority capitals have
schools and educated people, therefore, it is not surprising that there
are correlations of .72 and .62 between education and traditional authority
respectively, and both correlations are significant.

Later age marriages in the urban areas and the use of contraceptives
to a greater degree by women in the large towns are some of the factors for
urban-rural differentials in fertility in Ghana. The use of contraceptives
has been found to be more prevalent among the educated and urban dwellers.
The use of contraceptives is not a common phenomenon in SouthéMamprusi.

As one of the medical doctors in Nalerigu pointed out, it is even not an
accepted practice to the people who see no need for it. The only people
who have ever tried or used contraceptives are those who have received an
education. Extended formal education is also one reason for postponement
of marriage by some women in Ghana. In South-Mamprusi, education has not
yet made any marked impact on fertility in terms of reducing it. For
instance, the 1960 census shows that only 1% and 2% of the population has

past education and present education respectively.

(d) Another Locality: Among the migration variables only this one is of

 

any importance. It has already been noted that local mobility is more

important than inter—regional migration. Apart from traditional authority

77

as explained from Table 4, being an initiator of local mobility there

are other considerations such as a few families in search of new land

and migration of families to areas where relatives do live. The extended
family system has some measure of "social security" for members of the
society, such that it is not uncommon for young men or girls to be sent
by parents to relatives to live with them. A relative at any place and
facing economic or social hardships can migrate locally to any relative
who is likely to provide him with some sort of support. There are weak
correlations between the variable born in another locality and the other
variables, the highest being with foreign-born (r=.30). A few of the
foreigners, especially the Moshi, Busanga, Yoruba, Fulani, and Songhai,
engage in step-internal migration. Most of these groups are very mobile
in any region or between regions of Ghana. In South-Mamprusi they do
indulge in both intra and inter-local migration. What does not stand out
clearly and which is difficult to explain is the weak association between
the variables traditional authority and per cent born in another locality.
An examination of the raw data shows that apart from Nasia, a bridge

town which has 58.8 per cent born in another locality, some of the high
figures for the variable percent born in another locality are to be found
found at the traditional authority capitals--eg. Nalerigu 41.9%, Gambaga
26.2%, Nagbo 34.7%, and Kpasenkpe 33.3%. What might have happened here
was the existence of a threshold since both traditional and non-traditional
capitals have high figures. This suggests, as mentioned above, that
traditional authority alone cannot account for local population mobility.
Part of the explanation lies with other factors, such as the building of

roads, the search for new land, and the establishment of urban centers.

78

Even though the raw data shows urban centers with high figures for the
variable, per cent born in another locality, there is a weak correlation
between urbanization and born in another locality. The development of a
possible threshold is suggested again, for this result since some rural areas

also have high figures.

(e) Densityiof Population: It was assumed that there is a close rela-

 

tionship between density of population and growth of population. The
rationale behind this assumption was that where the people are many the
higher rates of population increase take place there. Even though density
of population has a simple correlation coefficient of .72 and is significant
at the .001 level, with population change it only explains about 1% of the
total variation in the stepwise regression model. As has always been the
short—coming of density analysis, the pockets of high population concentra-
tions, which in a few cases happen to be the urban centers, might have dis—
torted the importanceof this variable. Perhaps its relative unimportant
role can be explained by its close association with the other important
variables, namely traditional authority, urbanization,Education, with which
it has inter—correlations of .64, .79 and .54 respectively-—they are all

significant at the .01 level.

(f) Foreign Born: This is the last of the most significant variables.

 

Although it explains about 1% of the total variation it has .52 simple
correlation with population change. For a long time now there are quite

a few foreigners in South-Mamprusi as revealed by Table 10; namely, Yoruba,
Moshi, Mande, Songhai, Fulani, Hausa, other Africans and people from out-

side Africa. These,of,course have contributed towards population growth

79

in South—Mamprusi, especially if one takes into consideration the long
standing relationship between the area and other groups outside Ghana.

Most of the foreigners stay in the commercial and traditional authority
centers. Hence, between foreign born, and traditional authority and
urbanization there are correlation coefficients of .34 and .49 respectively,

which are significant at the .05 level.

(g) Sex Ratio: This has turned out to be one of the insignificant
variables and has a simple correlation r=.23 with population change. This
variable was included with the hope that it will serve as an indicator

of migration patterns, a component of population change. Sex ratio for

the population aged 15-44 years has a maximum value of 176 men per 100
women (Nalerigu) and a minimum value of 99 (Zangu). It has often been
argued that since men are more migratory for long distances than women

a low sex ratio of the 15-44 years age group is indicative of out—migration
and a high one is indicative of in-migration. With.these high sex ratio
figures, South Mamprusi is unlikely to fall into any of these categories for
as discussed previously little outwmigration is carried out and therefore
men stay at home. In like manner, the area has no significant economic
attractions for only male immigrants. The only exception is Nalerigu,

with the highest sex ratio, which had construction workers putting up the
hospital in 1960. The existence of a hospital, in addition to the fact
that it is the capital of the paramountcy, encourages young men to migrate
to it. The areas with low sex ratios, e.g., Zangu, which had —4% decrease
of population between 1948 and 1960 suffered greatly from the sleeping
sickness which had a high incidence among men. Other enumeration areas

with low sex ratios are Fio, Soo Saboya, Du, Nabari, and Buzulungu, the

80

uuoamm msmaoc coma msmzc “oopsom

NH magmas

'1 ll

 

223.551 .5 33:5

 

 

 

 

 

 

 

 

 

 

 

 

 

' “C‘ m‘
- .- ~*M- ,..~ v. —-—~— “w.-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-NmanuNPIoa—mmemmsm
.1 -1411“. m.
Mm J...
I. I
c. L.
-2
HS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 L cm...
W 1w. mm,” mwc<i

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ST. N
so In

mcq

 

 

m

 

 

comciafidfinlltltaqfigomfi

 

81

only places which decreased in population between 1948 and 1960 (-7,
—14, -25, -5, and -2 respectively). Indeed, records indicate that all
these places were camp centers for the eradication of sleeping sickness.

A population pyramid showing the composition of the population
by age and sex do conform broadly with what has taken place in an area
regarding fertility, mortality, and growth rates. Therefore Figure 12 is
a summary of what has been said about p0pu1ation growth in SoutheMamprusi
from the previous years to 1960. Considering SouthéMamprusi as a whole,
the sex ratio is very high with ages 0—10 years, and this is the largest
portion of the population. The large proportion of the young population
can be attributed to the lowering of infant mortality by medical care.
Caldwell has suggested that this could also be due to misstatements of ages
among ages 10-20 years, which has resulted in the 10—20 year age group
being relatively small.13 The ages between 20 and 40 years show a higher
number of females, and as mentioned previously, this can be explained by the
fact many males died owing to the sleeping sickness. The high sex ratio
of ages 40-60 is probably due to the occurrence of Cerebro Spinal Menengitis
in the 19203 which killed more women than men. The pyramid shows a few
aged people and this is a reflection on the low rate of life expectancy in

South-Mamprusi.

(h) The Distance Variables: It was hypothesized that there Should be

 

inverse relationships between population growth and distance from main
roads and distance from the nearest urban centers. There are correlations

of r=--60 and r=--51 between population change and distance from main roads

 

13Caldwell, John C., op. cit., 1967, pp. 30-35.

82

and urban centers respectively and both are significant at the .01 level.
This supports the hypotheses formulated. The role of urban centers has
already been discussed and since it was thought that facilities in urban
centers do encourage population growth, then the resultant correlation

is of significance in terms of urban-rural differentials. The building

of roads in South-Mamprusi attracted many people to settle along them.

The radial pattern of the distribution of population, especially along the
Walewale and Nalerigu road is clearly shown on the maps for the distribution
of population in 1948 and 1960. (Figures 13 and 14) These maps show the
growth in population during the period, especially at the urban centers.
Generally,the areas around Walewale in the West, and Nalerigu in the east
have dense populations.

There is an inter-correlation value of r=.85 between the distance
variables and the author has questioned whether one variable might not
serve the purpose. A problem in using only one is that not all the settle-
ments along the main roads are urban centers. Besides, the role of road
in attracting settlements along them and thereby raising densities and
consequently population growth rates, have radial spatial characteristics

while urban centers tend to result in a few clusters.

 

(1) Born in this Locality: This has a negative correlation value of
r=°-32 with population change and is significant at the .05 level. Its
role is very doubtful since it considers the main section of the population.
There are a few places which had high percentages born in this locality,

for example, Janga, 94.4%; Gbandaa, 86%; Bugiya, 85%; Du, 88%. One
explanation is that the per cent born in this locality does not, by itself,

tell the researcher the number of people that were added to the population.

83

 

A.uuomom wamooc coma momsuc

 

 

 

 

 

 

 

 

 

 

mmsmhun omxcommmxuc swnozum swauoamzu< "msoama>an
ma musmam
.nciu .nvgu
_ _ .090.
on} u. . .
on“ u. o o 0.
OO—ufl_o Au.
0 O 0 ”ca a \\.II
o O o VFIII'linl‘ol‘ o
o \ . o o
. o o s o . o
I. 00 O O \‘O o o O o O —
. o o o Ira—00
O I 00. O O o O \‘ O o o o m 0
O . O o o o 0 o o 000 0 o o 00 o
I . do; . ,2...
.. .o o o o o .
3m.
_ _ _ _
.900 .0m00 .nvoo .00 o—
mxoam2>an a<zcahan<ma rm zcah<acmcm mo zcahbmamaman

 

 

84

Auuomom mamsoc coma mcmscc

mwsmhun moxaowmmxlc swsszum swauoaozu< "mcoama>aa
«a enamam

 

  
  

 

 

    
  

   

.926 .930
.090. _ o8.— u. _
o? n 0
8. u . .
w
2.0. I l

huncnzfill

 

 

Coo.

 

 

_ _ a _
.900 .0n00 .nV00 .000.

 

 

 

 

monmH>Hn A<ZOHHHQ<MH Mm ZOHH<ADmOm mo ZOHHDmHmaman

85

The author has a feeling that this variable ought not to have been added
to the variables for the study. In view of its insignificant role in the
step-wise deletion programme one may be right to assume that it has not
distorted very much the results of the regression model. An interesting
significant inter-correlation, however, exists between foreign-born and
this variable (—.44). This explains the point that the foreigners do not
usually stay in rural areas; therefore, where there is a high proportion
of people born in a locality, there are a few or at times no foreign
residents. Earlier in the study, the contention was made that very few
foreigners settle in Mamprusi as subsistence farmers and this correlation
supports the contention. The per cent born in this locality has a very
insignificant correlation with education, r=.028, and shows the low level

of education in the vepy rural areas.

(j) Dependency and Fertility Ratios: These variables were used with the

 

hope that they will serve as fertility indicators, especially fertility
ratio. They have, however, turned out to have insignificant correlations
with population change (dependency, r=.18 and fertility, r=.36). Individua~
11y, they contribute very little to the total variation in the multiple
regression model. It was noted in Chapter II that Bogue and Palmore feel
that the fertility ratio underestimates fertility because the net result
of the ratio is biased. This could be a reason for the weak role of the
fertility ratio. In any case, it has inter—correlations of r=.43 and
r=.44 with traditional authority and urbanization respectively which are
significant at the .05 level. This supports the assumption that fertility
is higher at the urban centers and at the traditional authority capitals.
Though the dependency ratio is directly a function of the components of

population change in the sense that increase in population always implies

86
a higher dependency load than do stationary populations, its role here
might be due to the distribution of the ratio among the enumeration areas.
One reason might be that there was a development of a threshold in the
distribution of values of this variables between areas with high population
increases and areas with population decreases. On the whole, fertility
ratio with the correlation of r=.36 and significant at .05 with population

change, is more important than the dependency ratio.

(k) Born in Another Region: The minor role of this variable in accounting

 

for population change can be explained in terms of the under-developed
nature of Mamprusi at the national level. As Ravenstein pointed out, usually
"migrants proceeding long distances generally go by preference to one of the

- 1
great centers of commerce and industry." 4

Thus, a few migrants from far
away places within Ghana do come to Mamprusi. The few long distance migrants

in the area are in these occupations: business, administration, rural-

commerce, education, and fishing.

A Consideration of the Residuals:

From the regression analysis about 88 per cent of the total variation
of population change was accounted for by the thirteen independent variables.
(Table 12.) Both the delete and addition programmes yielded the following
variables as the most important - traditional authority, urbanization, educa-
tion, another locality, density, and foreign born; together they account
for about 87 per cent of the total variation of population change. For the
analysis of the residuals in view of the relative insignificant roles of the

other variables, only the above variables have been used, i.e. residuals

 

14Ravenstein, E. G., op. cit. (1885), p. 199.

87
from the step-wise deletion programme.

The enumeration areas have been divided into four categories of
population change namely, high growth, moderate growth, low growth and de-
crease (Appendix B). Figure 15 gives the negative and positive residuals
according to one standard deviation units. It has already been noted that
negative residuals indicate over-estimation of the per cent of population
change while positive residuals mean an under-estimation of the per cent
of population change. The pattern on Map 15 indicates that there were not
any striking deviations from the explained estimated variation, the highest
being only 2 standard deviation.

An examination of Figures 11 and 15 reveals the following: (a) All 71:5
the areas which had population decrease are in the negative one standard
deviation category, indicating an over estimation. (b) About half of the
areas with high population change (Langbinsi, Nagbo, Nalerigu and Gambaga)
have negative value of one standard deviation, while the remaining half are
in positive one standard deviation category. (c) In like manner, half of
the areas with moderate and low growths have negative one standard deviation
values while the rest are in the positive one standard deviation category.

(d) There is one extreme case that has a positive value of two standard
deviation; this is Gbeo with moderate population growth.

It is not surprising that the residuals indicate an over-estimation
for all the areas with decrease growth. The variables considered above are
insignificant in these places. The level of traditional authority is at the
village level, and apart from Du and Arigu which were considered as urban
centers at the lower level, the rest are rural areas. Education is also not
important in these places and densities are quite low. It was noted in

Chapter IV that these are the areas which did have high death rates in the

88

 

 

 

 

 

 

148008.. 20_hu..uo no.3..nuhm

 

muswam
« 2 5.. .
:02!
r D D h
F o. o . ‘ C ‘ J
7 0. 0

ml.— 508$

ma<=n.mu¢

 

 

89

past. Thus independent variables on the vital statistics of the area --
births, deaths and migration -- would have helped, however these are not
available.

The position of Langbinsi, Nalerigu, Nagbo, Gambaga and Gbimsi
in possessing negative residuals though they have high growths is not very
clear. All these places have some level of traditional authority, urbaniza-
tion, education and quite a number of foreigners. Being traditional and
urban centers they will also have immigrants from other parts of South-
Mamprusi. If this is the case and yet the residuals suggest an over-
estimation, perhaps too much importance was attached either to a few or all
the variables. One shortcoming of traditional authority is that there are
only a few places with Paramount chief (Nalerigu) and Divisional chiefs.
This limited spatial distribution might have had some distortion on the
effectiveness of the other variables. There is also the point that with-
the establishment of the Indirect Rule by the British administration during
the colonial period the importance of traditional authority capitals was
reinforced. They became administrative, health, commercial educational and
transport centers. Thus they do possess all the qualities of the most
important independent variables.

Nasia, Gambaga, Gbandaa, Burugu, Tangbini, Wulugu, walewale and
Guabuliga are all high growth areas and yet the residuals indicate that
there was a little under-estimation of the per cent change of population.
Gambaga, Walewale and Wulugu are urban centers but the others are rural
areas. Almost all these places are near main roads. If the values have
been under—estimated one could infer that the areas might have benefited
from the use of other variables. What has been said about the need for

vital statistics is very true for these places. The fact that they are

90

near main roads do suggest continued receipt of immigrants therefore a
net-migration variable would have been useful, but such information is
not available.

The areas with moderate and low growth rates can broadly be
divided into under-estimated and over-estimated areas. (Appendix 2)
Most of these are rural areas but a few like Sakogu, Kpasenkpe, Wungu,
Bong-Da, Janga, Kparigu are both traditional centers with some measure of ‘7‘
education and urbanization. Wungu, Kpasenkpe and Janga have divisional
chiefs and yet they all indicate an over-estimation of the per cent change.

This again suggests that though cultural considerations are important for

 

population growth (in view of the minor deviations) there are also other -~'
factors which should be given, some attention. It could also be a pointer
to the already mentioned uneven distribution of the major traditional
capitals, for in the West only Janga, Kpasenkpe and Wungu are divisional
capitals.

The extreme case of an under-estimation in Gbeo with a.+2 standard
deviation value, can be explained in terms of the density variable. Though
the area has moderate growth it has a far lower density of population than
other areas with moderate growth, owing to the extensive area it occupies.
Kurugu was mentioned in Chapter IV as a former divisional capital but
which lost its position in the 19303. This resulted in an exodus of quite
a few people from the area hence the low density of population today. Other
variables based on vital statistics might have helped in explaing the varia—
tion here.

A major striking feature of Figure 15 is the extensive belt of

under-estimated values (+1 Standard deviation values) almost in the middle

91

of the area, from east to west. Indeed, there would have been a contiguous
belt if Wungu, Gbimsi, Langbinsi and Bugiya had the same values. In an
attempt to explain the pattern the author realized that there are the
following factors about this middle belt of underestimated per cent popula-
tion change. (a) It contains most of the oldest settlements in SouthéMamprusi.
(b) It has some of the most important food producing villages in Mamprusi.
(c) The eastern part of the belt is along the Walewale - Nalerigu main road.
(d) A superimposition of the residuals on a relief map (Figure 2) indicates
that a majority of the settlements in this belt are situated along the head-
waters of the streams of the White Volta and River Nasia. A consideration
of all these factors led the author to believe that there ought to have —'
been an "agriculture land value" or "agricultural production" variable for
the regression analysis.

In choosing the sites for the earliest settlements the availability
of water and fertile land were of paramount importance to the people.15
The choice of the area east of Walewale by the Tampolensi because of its
fertile soil has already been noted. All these villages lie within this belt
and Langbinsi with a high population growth is not only a food market for
South—Mamprusi, but also for North—Mamprusi and some parts of Dagomba. Another
consideration is the spatial distribution of immigrants from North-Mamprusi.
It was noted earlier on that most of them are agricultural settlers and are
therefore to be found within this underestimated per cent change belt. In
view of this, the author wonders if a variable on net migration might not
have reduced the striking feature of underestimation in the belt. Lack of

any data on agriculture land value, agriculture production or rural incomes

 

15Nabila, John S., op. cit. (1968), p. 9.

92

and net-migration made the inclusion of such variables not feasible.

B. Post—Censal Period - 1960-1969

Basically the main objective of the sample survey was to
collect data on the settlement and population geography of SouthrMamprusi.
Thus in addition to providing answers for the questionnaires the author col—
lected data on other related conditions. In the course of processing the
data collected in the field it was considered necessary to focus the
thesis on only population change in Central SoutheMamprusi. Thus the

following analysis will be based on the major components of population

change.

1. Fertility and Mortality: Total fertility rates were computed for the
enumeration areas. This rate is the estimate of the number of children a
cohort of 1000 women would bear if they all went through their reproductive
years. It is computed by summing the age-specific fertility rates for all
ages and multiplying by the interval into the ages grouped. The age-specific
fertility rate is the ratio of births by age of mother to women in each

age interval.

According to Bogue, the total fertility rate is the best single
cross-sectional measure of fertility because it is rather closely restricted
to the child bearing population and is not influenced by differences in the
age composition between child—bearing populations. It is also independent
of mortality since it assumes that all women survive from birth to the end
of child—bearing period.16 To Caldwell, one measure of fertility which has

the advantage of summarizing a considerable amount of information in the

16Bogue, Donald J., op. cit., 1969, p. 659.

 

93

form of a simple index is the total fertility rate. He also indicates
that fertility levels in Ghana have not changed much, therefore the total
fertility can be equated with the average size of completed family sizes.17

Information on current fertility, i.e. births occurring during
the twelve months preceding the survey was obtained from the field. In
addition, the total number of children ever born to women and the number
of children alive were also recorded. This information was used for the ‘11
computation of the total fertility rates. More often than not, the births
occurring to women aged 50 years and above are regarded as the total fer-
tility. It was not possible to compute the crude birth rate since the
mid-year population was not known.

The table below gives the mean number of children ever born to

women by age group in South-Mamprusi. The completed fertility rate is

TABLE 15

MEAN NUMBER OF CHILDREN EVER
BORN TO WOMEN BY AGE GROUP

 

 

Age Group mean No. of Children
1. 15 - 20 1.0
2. 21 - 25 2.0
3. 26 - 30 3.2
4. 31 — 35 4.3
5. 36 - 40 5.1
6. 41 - 45 5.8
7. 46 — 50 6.0
8. 51 — 55 6.5
9. 56 - 60 7.0
10. 61 - 65 7.0
11. Over 65 7.0

 

 

l
7Caldwell, op. cit., 1967, p. 88.

94

TABLE 16

TOTAL FERTILITY AND THE NUMBER OF

DEATHS FOR THE PAST TWO YEARS

 

Enumeration
Areas Visited

Total No. of deaths for the
Fertility past two years

 

 

1. 800 Daboya 5.9 26
2. Nasia 6.0 15
*3. Kparigu 6.5 20
4. Zambulikura 6.8 25
*5. Bong—Da 6.1 15
6. Kpikparibobgu 5.9 22
7. Burugu 7.2 35
8. Samene 6.3 26
9. Guabuliga 6.4 19
10. Zangum 5.9 22
ll. Logri No. 2 6.3 16
12. Sayo 7.0 20
*13. Kpasenkpe 6.9 17
14. Kinkadina 5.8 28
15. Gbeo 6.0 30
16. Kpakpiri gbangu 5.9 16
17. Namasim 6.1 17
*18. Sakogu 6.5 18
*19. Janga 6.2 20
*20. Nalerigu 7.1 10
*21. Wulugu 6.8 15
*22. Walewale 7.3 15
*23. Gambaga 6.2 17
*Traditional and Urban Centers.
seven. Though this is the average, there were a few completed family

sizes which were about ten. This was largely due to past infant mortality

rates which lessened the spacing period between births. Considering South-

Mamprusi as a whole, the average length of the interval between births is

2 years but this was noticed to be shorter whenever a child died before the

weaning period.

It is most likely that these figures are lower than what

exists because many people are reluctant to discuss their dead children.

95

Table 16 gives the total fertility rates and the number of deaths
for the past two years preceding the sample survey. The fertility rates
indicate that South-Mamprusi has a minimum total fertility rate of 5.8
and a maximum of 7.3 for Walewale, an urban center. This is not very
different from what Gaisie found out for the area and this shows that
fertility rates have not changed very much.18 The slight increase might
be due to increased medical care, and on the whole better living conditions.
The total fertility figures are quite high in a few of the traditional
authority capitals and urban centers. This still suggests that fertility
is higher in these areas than in the typical rural areas. There is no
doubt that as urbanization and education continue to develop in South?
Mamprusi, these fertility rates in urban centers will drop. Though most
of the respondents felt that there was nothing wrong in having large
families, they expressed concern about the changing cost of living which
makes it increasingly difficult for large families to be catered for.

The crude death rates could not be computed for the enumeration
areas because there was no idea about the mid-year population. The number
of deaths for the past two years, however, helped in giving an idea about
mortality levels in these enumeration areas. From Table 16 it is quite
obvious that deaths do occur often in the rural areas than in the urban
centers. On the average, the people reported that deaths in recent years
have been less than what they were years ago. This is definitely due to
the availability of medical facilities especially in areas near the main
roads, urban and traditional authority capitals. Besides, since the

eradication of the sleeping sickness epidemic in the 19308 no epidemic

 

18Gaisie, S. K., op. cit., 1969, p. 39.

 

96

has taken place in South—Mamprusi. The major causes of deaths among
the young population were malaria and measles. These are somewhat periodic
diseases every year, however they are being controlled and deaths from them

now are lower than what the situation was years ago.

2. Internal Migration: This seems to have become very pronounced during

 

the period 1960-1969. The search for better agriculture lands by a few
people in the rural areas continues and this often results in the move?

ment of single families or complete villages. Along the Gambaga -Walewale
road new settlements are continually being established by people who move

to that area from other parts of Mamprusi, especially, the section around
Gambaga. Namiyala, Yaroyiri and Tianoba are all new settlements on the

road and were established between 1960 and 1969. Namiyala, for example,

was established by people from a whole village near present Zagiri in the
South—east. These villages along the road have the advantage of utilizing
some of the fertile areas in Mamprusi. Besides, With.the road it is quite
easy to send food to be sold in markets in Walewale, Langbinsi, Gambaga

and Nalerigu. A few settlements are also being established along the other
main raods, for example between Wulugu and Nabari or between Nalerigu and Sakogu.
Bogue has pointed out that internal migration is the single factor accounting
for population growth differentials in many places in the world. At the

rate that families are increasingly becoming mobile in SouthfiMamprusi it

will be internal migration which will account for population growth differen—
tials in the future. At present it is only one of the factors accounting

for population growth differentials.

Immigration: This continued to be important in SouthrMamprusi.as it was

 

in the previous years. Immigrants continued to migrate into the area from

J‘fl

 

97

North-Mamprusi and Central MOshi.
In the rural areas immigration was found to be unimportant. Of
the households visited in the rural areas only 15 families were immigrants
who had moved into South-Mamprusi between 1960 and 1969. Apart from some
immigrant fishermen in Kpasenkpe and Janga, the immigrants were agricultural
settlers. The fishermen were either Battor from Central Togo or Hausa,
but the agricultural setters were from Frafra, Kusasi and Moshi. None of 'l-W
these regarded South-Mamprusi as their home, their real homes were back 1
I
on their father's land in the various places they migrated from. The
Frafra and Kusasi made periodic visits home to take part in the major annual
festivals.
The urban and traditional centers received a majority of the
immigrants. Walewale is a commercial center for many places in the West
and therefore received most of the immigrants to the West. These were some
Yoruba, Hausa, Moshi, Dagomba and people from Southern Ghana, who moved
into South Mamprusi in order to engage in trade. Another source of
immigrants was through administrative functions. Walewale, with a health
center, Workers Brigade Camp, Post Office, Agriculture Extension Station
and a Middle School received quite a substantial number of both educated
and uneducated workers. Other places in the West which.received immigrants
were Wulugu, Gbimsi, Wungu, Nasia, and Logri No. 2. At the time of the
survey the author estimated that the population of Walewale was well above
5,000, perhaps about 6,000 people.
Nalerigu has become an important destination for more than half
of the immigrants to the east. Indeed, the major component of population
growth here between 1960-69 should be immigration, both.from outside Mamprusi

and from other localities. In addition to its role as the seat of the

98

paramountcy, it is the administrative headquarters of the Nalerigu

Local Council and a medical center. The growing importance of the hospital
in Nalerigu has brought new functions to the town such as construction,
commerce, building of houses for renting out to visitors and workers and

a fooddmarket.

The role of traditional authority continues to make Nalerigu the

.J

termini of some chiefs and their families. Nalerigu exhibits some

—'—_ - a t.‘__£_

features of stream and counterustream in local migration. For as chiefs
attain higher ranks and qualify to stay in Nalerigu, young princes become

lower chiefs and move out to other villages. At the time of the survey

 

there was a new paramount chief -— Na Bongu —- who had been installed

in 1968 to replace a previous one who died in 1967. Na Bongu was the
chief of Kparigu and therefore had to move to Nalerigu with all members of
his family, his patri- and matri—kins, his close elders in the court and
other members of his extended family. Since he became a chief about

three princes have moved out of Nalerigu to be chiefs in other places.

The Nalerigu Hospital itself has established a village for the
treatment of tuberculosis patients. These patients are from different
places in Mamprusi and outside the district. They stay here until they
are out of the danger zone or completely healed. At the time of the survey
the author enumerated fifteen T. B. patients with their families. The
average number of years spent in the village was two years, and while they
are in Nalerigu they are given land to farm in order to support their fami-
lies. Apart from this village, the hospital has many employees including
American and British medical officers and their families. Further, the
hospital also draws quite a high number of patients every day from other

places in Northern Ghana and a few patients from the South. Nalerigu should

99

have more than 5,000 people now.

In the east, Sakogu, Gambaga, Nagbo and Langbinsi, were some
of the places which received migrants from outside Mamprusi. Immigrants
from Bimoba and Komkomba areas in the east are among the people who move
into the east to settle. They are usually agricultural settlers rather

than people engaged in commerce.

4. Out-migration: This is becoming very important, as Table 17 indicates

 

in South-Mamprusi than the situation was years ago. The table shows that
quite a few people engaged in long distance migration. All these were
mostly men who went to Southern Ghana in search of better income or in
order to have the experience of staying in the relatively better developed
southern part of the country. Ashanti and Bong-Ahafo regions are the

usual destinations for these migrants who were mostly in the 14933 years
age group. Few of them stay in urban centers in the South while a majority
of them work on cocoa farms on contract basis, for two or three years.

When the contract expires, a migrant can renew it or return home.

Among the out-migrants enumerated were some females but these
usually went with their husbands or joined them later on. However, the
common practice is for a long—distance migrant male to leave the wife
and children at home to be taken care of by other members of the extended
family.

Education was found to be one of the factors encouraging out-
migration. Many students from SoutheMamprusi are in other parts of Ghana
receiving education in secondary schools, colleges and universities.

Young men and girls with past education who could not make it to the end,

especially the young men, engage very muCh in out-migration to Southern Ghana.

 

100

TABLE 17

LONG DISTANCE MIGRATION
IN SOUTH-MAMPRUS I

 

 

No. of Average
Enumeration long distance number of
Area emigrants years away

l. 800 Daboya 6 2
2. Nasia 6 3
*3. Kparigu 15 4
4. Zambulikura 5 3
*5. Bong-Da 4 1
6. Kpikparibobgu 20 5
7. Burugu 10 2
8. Samene 18 4
9. Guabuliga 23 3
10. Zangum 30 3
ll. Logri No. 2 15 2
12. Sayo 40 6
*13. Kpasenkpe 20 3
l4. Kinkadina 15 3
15. Gbeo 25 8
l6. Kpakpiri gbangu 30 6
l7. Namasim l4 2
*18. Sakogu 50 8
*19. Janga 25 5
*20. Nalerigu 50 4
*21. Wulugu 100 8
*22. Walewale 60 3
*23. Gambaga 40 5

 

*Traditional Authority and Urban Centers.

The number of outdmigrants reflects the structure of the population
of the various enumeration areas as well as the economic opportunities avail-
able in them.

The number of out—migrants from rural areas is lower than that from
urban centers. This suggests that there is still an attachment to the "family

system" in rural areas than the case is in urban areas.

ULv._

 

101
A closer examination of Table 17 shows that some rural areas

have many out-migrants. These are the enumeration areas with immigrants
from North—Mamprusi and Moshi who usually engage in step-migration to the
South, with Mamprusi as one of the intervening destinations. Sayo, Gbeo,
Kpakpiribangu and Sakogu have all got high numbers because of these immi-
grants who tend to be very mobile. wulugu has the highest number - 100 -
because of its location. It is on the main Bolgatanga - Knmasi road and
it is now an aspired after goal by the youths to migrate to the South.to
stay for a few years before returning home. Indeed, there is no house

in Wulugu without some young men, either alone or with.their spouses, living

in Southern Ghana.

CHAPTER VI

SUMMARY AND CONCLUSIONS

This has been a consideration of population change in Central
South Mamprusi. An attempt has been made to find out the factors account-
ing for population change in the area as a whole.

For the purpose of the study, the following assumptions (hypotheses)
were made:

(1) Reproductive change (natural increase) is the major source
of population change in the area under study. The assumption here was
that, since South-Mamprusi receives a few immigrants it should be repro—
ductive change rather than migration that is accounting for population growth.

(2) There is a close relationship between traditional authority
and growth of population. For a very long time now a hierarchy of chief—
doms has been developed such that the growth of some settlements is due
to the existence of the chieftaincies. It was hoped that, since these
centers serve as capitals for the other villages they should show higher
growth rates.

(3) Men migrate for long distances more than women and therefore
the sex-ratio of the age group 15-44 years is a good measure of migration.
More often than not, in the developing countries, men migrate for long
distances to the "pull" areas (economic cells). This creates a high.sex
ratio in the "pull" areas while there is a low sex ratio in the "push"

areas. Thus this variable was to find out the level of in and out migration

102

103

in South—Mamprusi.

(4) There is a direct relationship between density of population
and growth of population.

(5) There is an inverse relationship between density of population
and distance from main roads. Thus, there is also an inverse relationship
between population growth and distance from main roads. The assumption
here was that settlements along the roads should grow faster than those
which are far away.

(6) The urban centers are areas of rapid population growth, thus
there should be a direct relationship betwen population growth.and urbanizav
tion. By the very nature of their functions there is an inverse relationship
between population growth and distance from urban centers.

(7) There is a close relationship between education and the ten-
dency to migrate. There should therefore be a direct relationship between
education and population growth.

The changing patterns of the increase in numbers of population
in the twentieth century was considered in the study. For the early period,
population change was accounted for at the local council aggregate level,
that is, considering Central SoutheMamprusi together. For the period between
1948 and 1960 the study was based on the enumeration areas of the 1960 census
and the author's field work. Data for the analysis as a whole was derived
from National Census Reports, records in the National Archives and data from
the author's field work. Since the 1948 and 1960 census reports are more

reliable than the previous censuses, it was considered quite helpful to

consider population change in the enumeration areas at the distributive level.

In order to understand the mechanisms for population change in

South-Mamprusi, fourteen variables were used for carrying out a multiple

_-—-—~‘ '—- '- R11

 

‘7

104
regression analysis. Population change was the dependent variable while
the other thirteen, derived from the above hypotheses were the independent
variables. (Refer to Chapter 2)

The study area forms part of the Voltaic family and linguistically
belongs to the Mole-Dagbani association, which covers a major part of
Northern Ghana and Upper Volta. It is known to have been settled for a —
very long time, perhaps before the neolithic period. One of the major F J}
aspects of the human geography of the area is the peopling of it by small

groups of people who immigrated into it to settle, beginning in about the

15th century.

 

Through these immigrants from the north, a system of traditional
authority was develOped and today there is a hierarchy of chiefdoms. The
small groups of people who moved in and introduced the idea of chieftaincy
and territorial acquisition form the royals today. This factor has been of
paramount importance in considerations of the general development of the
area as well as the patterns of population distribution.

Though the earlier censuses were not very accurate, they do indicate
that population has been increasing in South-Mamprusi since 1900. At the
turn of this century, fertility rates were supposed to be very high. In like
manner, death rates were also quite high, in veiw of the poor living con-
ditions at the time. The occurrences of the influencza epidemic in 1918
and cerebro spinal menengitis in 1921 raised the death rates. It has been
recorded that the high sex ratio in the 1920's was largely due to the death
of many females as a result of these diseases.

There has been a general desire for children to help in the sub-
sistence economy and this has consequently raised the fertility level.

Indeed, the customs and religion of the people do favor high fertility.

105

The creation of Gambaga as the first headquarters of the Northern
Territories encouraged a few people to migrate into the area. When the
capital was removed to Tamale the area continued to receive immigrants from
further north, especially from NorthéMamprusi and Central Moshi. South-
Mamprusi is very fortunate to have abundant land, about eight per cent is
cropped in any one year. The relative nearness of the area to the populous
North-Mamprusi, where by the 19308 there was a problem of over population,
has been a factor in attracting immigrants from these hard-pressed areas,
though not in considerable numbers as the administration anticipated. In
rural Ghana as a whole, many farmers find it difficult to leave their
"father's land" and this explains why many people from.NorthrMamprusi.have
not moved into SoutheMamprusi.

By 1948, health conditions of the people had improved and the
mortality rates would have been lower than those of previous years if it
had not been due to the occurrence of sleeping sickness in some villages.
The disease claimed the lives of quite a few people and the situation was
made worse by the invasions of locusts in the same period, which.resu1ted
in periodic famines. During this time South—Mamprusi continued to receive
immigrants from.North-Mamprusi, Central Moshi, Nigeria and Southern Ghana.

Birth rates continued to be the same as they were years ago. Death
rates, since the eradication of some of the major diseases coupled With.the
building of a hospital in Nalerigu and a health center in walewale, have been
on a steady decline. Infant mortality rates which were quite high during
the past years have been considerably reduced. Between 1948 and 1960 the
whole population increased by 54 per cent, this representing an intercensal
growth rate of about 3 per cent per year. Even though there were no major

urban centers by 1960, there were quite a few settlements which.were considered

 

106

as more important than the rural villages. These continued to grow
at the expense of the other villages and by virtue of the facilities
available in these urban centers they rendered some services to the rural
areas.

During the post-censal period population continued to increase
until the departure of most of the foreigners.1 At present, it is not
known exactly the effect this mass departure of foreigners has on p0pu1ation
numbers in the various enumeration areas. A sample survey revealed that
fertility levels were still the same and the family system was still geared
towards the possession of large families. The only exception being the
few educated people who have started possessing small families. Out~migration
is increasingly becoming important in SouthéMamprusi and it is mostly the

young men who emigrate to Southern Ghana to stay for short periods.

Hypotheses Accepted: Through the multiple regression analysis the author

 

was able to find out some of the major factors which influence population
growth in the study area:

(1) Traditional authority was found to be of importance for p0pu—
1ation growth. The hypothesis that it should have a direct relationship
with population growth was supported by the results of the regression model.

Before the establishment of the British Administration, the system
of chieftaincy was already in existence. The introduction of Indirect Rule
whereby the Colonial Administration did almost everthing through the chiefs

reinforced the importance of traditional authority. The traditional

 

1In November 1969, the Government of Ghana passed a new immigration
law requiring that aliens of Ghana should be in possession of valid resident
documents. This resulted in the departure of about three quarters of the
aliens, especially those from nearby countries.

 

107
authority capitals, therefore, became administrative, commerical educa-
tional, health, transport and religious centers. A few of them developed
from rural to urban status such that they have continued to receive
immigrants from other parts of Mamprusi and from outside the area. Indeed,
almost all of them are attracting areas for foreigners. The immigration

of Muslims into SoutheMamprusi and consequently the establishment of that

religion in the area has helped in raising fertility rates. At the beginning 57‘

of the study it was felt that a religious variable should be used, but

this was dropped due to lack of sufficient data. .
(2) Density of population was found to have a direct relationship é

 

with population growth, thus the hypothesis was supported at the .001 level.

(3) It was hypothesized that there should be inverse relationships
between population growth and distances from main roads and urban centers.

The roads continue to attract people to settle along them such that during
the field work relatively new settlements were found along these roads. This
hypothesis was supported by the results of the analysis.

(4) The hypothesis that urban centers grow faster in population than
rural areas was accepted. In many other places in Ghana urban—rural differen-
tials in fertility rates show that urban centers have low fertility rates.
However, in South-Mamprusi it was found that, owing to the fact that urbani-
zation on the same scale as is found in Southern Ghana is still taking
place, fertility was rather high in urban centers. The many facilities
in urban centers do tend to favor the prevalence of higher growth rates of
population.

(5) Education as a positive factor of population growth was accepted.
The acquisition of education does encourage people in Ghana to be more mobile

than the people who are uneducated. Since all the elementary schools are

108

day schools which cater for the local population, only people aged more
than 15 years with present or past education were considered. People

with past education were those who were in the area and engaged in adminis—
tration and other occupations. People with present education were students
from the local area attending school in other places, but were currently

at home, for the 1960 census was carried out during vacation. The only
exception will be localities with middle schools, in which case the students

were attending school in the same locality.

Hypotheses Rejected: (l) The assumption that reproductive change is the
major contributor of population growth was not accepted. A pitfall of this
hypothesis is that there was no specific independent variable in the study
based on reproductive change. The most appropriate variables for it would
have been those based on vital statistics.

(2) The contention that the sex ratio should be a good indicator
of the nature of migration was rejected. This variable has been found to
be significant at the national level as an indicator of migration patterns.
In South-Mamprusi long distance migration was not important hence its
insignificant role in the regression model.

On the whole, the variables used for the regression analysis, especi-
ally those of traditional authority, urbanization, education, born in another
locality, density of population, and foreign born, were significant for the
study. Population change was accounted for by 88 per cent by all the indev
pendent variables and this indicates the significant relationship between
these variables and population change.

Though this is the case, the availability of vital statistics would

have saved the author from a great deal of trouble from using other variables.

 

109

Indeed, such data would have helped in focusing the attention of this
study on the major components of population change. Though, the results
are quite good it is difficult to give the specific contributions of
the major components of population change, namely fertility, death and
migration rates.

The use of the variables pertaining to place of birth has a short
coming of not giving the researcher the opportunity of knowing exactly
the contribution of immigration. The variables do not take into considera—
tion the period being studied, such as distinguishing between old and new
immigrants. It was envisaged that the results of the multiple regression
model should serve as a reflection of the whole trend of population change
in South—Mamprusi. Hence, though the birth place variables did not give
an idea of conditions between 1948 and 1960, they gave an idea of trends
of population growth, regarding the contribution of migration, for the

whole period considered in the study.

 

110

GLOSSARY

l. Dagbansabilisi: Literally this means inferior Dagomba. Binger was
told that traders and all other people of Mande origin were called
Dagomba. From the name coupled with Binger's work it could be
inferred that the invaders and their kinsmen gave the original

people this name. The original people are at times called Tindanbisi,
meaning descendants of the Tendana.

2. Tendana: Before the 16th Century, i.e. before the chieftaincy
system was introduced,every village was an autonomous unit and the
head was the Tendana -- the OWner of the land. He was the represen-
tative of the 'Earth-God' and was consulted in matters related to
farming, hunting and spiritual affairs. After the establishment of
the chieftaincy system, the Tendana continued to play that role but
subordinate to a village chief.

3. Mamprugu: The following words have been coined from this name:
(a) Mamprugu - the kingdom
(b) Mamprusi — the ethnic group or the people
(c) Mampruli - the dialect spoken by Mamprusi

(d) Mamprugunaba - the Paramount Chief of the
Mamprusi.

111

APPENDIX A(l): DEPENDENT AND
INDEPENDENT VARIABLES

 

 

Variables Type . Symbol
Population Change Dependent Xl
Density of Population Independent X2
Fertility Ratio Independent X3
% Born in this locality Independent X4
% Born in another locality Independent X5
% Born in another region Independent x6
% Foreign Born Independent x7
Sex Ratio Independent X8
Traditional Authority Independent I X)
Distance from urban centers Independent X10
Distance from main roads Independent X11
Urbanization Independent X12
Dependency Ratio Independent X13

Population with past or
present education Independent X14

 

(Also refer to page 26 for list of variables).

112

 

 

\‘l‘
H

 

oo.a so. we. no.. ~o.u No. oN. mm. as. NH.I mo. on. so. os. x
oo.a Ho. oo.- Ha.u oa. oo. Ho.- ~m.u as. HH.I oa.n NH. oa. max

oo.H 55.- mo.. so. mN. as. so. no. o~.u ss. so. so. was

oo.a mo.- ms.- om.n om.- oo.u so. so. s~.u oo.- am.u Has

oo.a No. mm.. mm.- Ha.u Ho.- ma. o~.- o~.- oo.- oax

oo.H HN. so. no.. oa. sa.: ms. so. No. ox

oo.H, oo. mo. oH.- Hm. oo.- om. mm. ox

oo.H oH. om. ss.u so. os. an. ax

oo.H Ho. Ho. o~.- HH.I oH.- ox

oo.a no.. so.- mo.- mm. as

oo.H NH.: so. No.- sx

oo.H as. on. ox

oo.H Na. Na

oo.H ax
sax max use use ode ox ox ax ox ox sx me «x as -MHMM

 

mzcaH<ammmoo mamZHm mo NHMH<Z

Amv< Naczmmm<

113
APPENDIX B

ENUMERATION AREAS WITH PER CENT POPULATION
CHANGE AND RESIDUALS

 

 

 

 

 

 

Per Cent
Enumeration P0pulation One Std.
Area Change Residuals Deviation Category

1. Walewale 139 27.13 1

2. Langbinsi 127 -ll.66 -1

3. Nagbo 124 - 1.05 -l

4. Nalerigu 112 - 9.63 -l

5. Burugu 109 15.55 1

6. Tangbini 100 7.49 1 HIGH
7. Guabuliga 99 9.97 1 GROWTH
8. Nasia 89 8.76 1

9. Gambaga 89 -20.91 -1
10. Gbandaa 88 21.40 1
11. Wulugu 83 18.33 1
12. Gbimsi 80 - 4.01 -l
13. Yama 72 28.58 1

14. Gbeo 61 54.05 2
15. Wundua 56 -13.56 -1
16. Kparigu 44 10.32 1 MODERATE
17. Gbani 37 10.07 1 GROWTH
18. Kpasenkpe 35 -l4.11 -l
19. Wungu 32 - 9.73 -l
20. Gbango 31 26.70 -1
21. Sakogu 20 16.54 -1
22. Bugiya 17 -14.24 -1
23. Bong-Da 15 17.54 1 LOW
24. Mushio 11 2.26 1 GROWTH
25. Zambulkura 10 6.47 l

26. Janga 3 22.68 -1
27. Buzulungu - 2 31.45 -1

28. Zangu - 4 5.73 1

29. Nabari - 5 -14.47 -1

30. Fio - 7 - 0.66 -l DECREASE
31. Arigu -12 - 9.42 -l

32. 800 Daboya -14 -10.70 -1

33. Duu -25 -12.03 -1

 

 

 

ouum

 

 

 

114

 

.ouo ,wouon3 moon amnudoa woao: 30mm mamnB onumum mouosa «waaow wwsaoo soawom a wow ow< oaoz
A

.moou mao>mue lusm Na umoa um o: moo a umsz amuaumz waoa How mnm\o£ owmaaa>
Iauu<v ou soon 50% was Iooaxm momma 36m sommmm ma umsz oa muosz
manna so» on one? mesa oosz luaaom

Imm OH 053

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ZOHH<MUHZ um zmom

 

 

 

 

 

00am

mxumaom moauaosoo oooamuuo coauoooooo osumum oouoaanu mo>a3 ooauumz ow< xom uwuam
amoammnm Annmac umoawamc you no .oz mo Hoaso oaon
coaooooom condom .oz twosomv mamz

 

 

 

 

 

 

 

 

 

 

 

 

Aoaooomooo soc no.2moa
mosaozzoaammoo

no Naflmem<

 

ooum

 

 

 

 

mxumamm

mo whom soumonc
ucoom oommooon muonz

\omaa

 

comma so «mono

 

«mao on cam muonB

 

echo on oao zoos

xom

 

 

mm<

mamz

 

 

AoaO£mwnom hm mhmww N umma mnumwa mo .ozv NHHA<HMQZ "Q Zach

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.oum
5
1
l

mxuma Honuoz Hocumm Honuoz Hosumm Hoaumm mamaom Honumm am>uouaa onuooz nuuam annoy z Amvumow Na

lam mo ow< ou case no owouo>¢ Na unma aaaum mamaom
soaumonooo coaumosom Icoaumaom Amvoamz Susan nouoaaeb mo osmz
awaAaHmmm no zmcm
mmmHozoneomoo

U xaozmmm<

 

 

BIBLIOGRAPHY

Barclay, G. Techniques of Population Analysis. London and New York, 1958.

Berry, Brian J. L. and Duane F. Marble Spatial Analysis: A Reader in
Statistical Geography. Prentice-Hall, New Jersey, 1968.

Blalock, Hubert M. Social Statistics. McGraw-Hill Book Company, New
York, 1960.

Binger, L. G. Du Niger au Golf de Guinee. Vol. 2, Paris, 1892.

Boateng, E. A. A Geography of Ghana. Cambridge, 1966.

Bogue, Donald J. Principles of Demography. John Wiley and Sons, 1969.

. and James Palmore "Some Empirical and Analytical Relations

Among Demographic Fertility Measures with Regression Models for
Fertility Estimation," Demography, Vol. 1, 1964.

 

 

. and Dorothy L. Harris Comparative Population and Urban
Research Via Multiple Regression and Covariance Analysis. Scripps
Foundation, 1954.

 

Busia, K. A. "Some Aspects of the Relation of Social Conditions to Human

Fertility in the Gold Coast," in Lorimer, Culture and Human Fertility,
Paris, 1954.

Caldwell, J. C. "Population Change" in Walter Birmingham, 1. Neustadt
and E. N. Omaboe (eds.) A Study ovaontemporary Ghana, Vol. II.
Some Aspects of Social Structure. London, 1967.

. "Fertility Differentials as Evidence of Incipient Fertility
Decline in a Developing Country — The Case of Ghana," Population
Studies. Vol. XXI, No. 1, July, 1957.

 

 

. Fertility Attitudes in Three Economically Contrasting
Rural Regions of Ghana," Economic Development and Cultural Chang_.
Vol. XV, pp. 217-38.

 

 

. African Rural-Urban Migration: The Movements to Ghana's
Towns. Canberra, 1969.

Clarke, John 1. Population Geography, Pergamon Press, London, 1965.

116

117

Coale, Ansley J. "How a Population Ages or Grows Younger," in Ronald
Freedman, (ed.) Populatiop: The Vital Revolutipn, New York:
Doubleday and Co., Inc., 1964.

 

Coontz, S. E. Population Theories and Their Economic Interpretation. 1957.

 

Demko, George J., Harold M. Rose and George A. Schnell. Papulation Geo-
graphy: A Reader. McGraw-Hill Book Company, New York, 1970.

 

Dickson, K. B. A Historical Gepgraphy of Ghana. Cambridge, 1969.

 

District Diaries, Gambaga Administrative Office, Gambaga.

 

Fortes, M. and Evans Pritchard. African Political Systems. Oxford Univer-
sity Press, London, 1940, p. 7.

 

Fortes, M. The Web of Kinship Among the Tallensi. London, 1949, p. 83.

Forde, Enid R. The Population of Ghana: A Study of the Spatial Relation-
ships of its Socio-Cultural and Economic Characteristics Studies
in Geography No.15. Northewestern University Press, Evanston,
Illinois, 1968. -

 

 

 

Gaisie, S. K. Dynamics of Population Growth.in Ghana. Legon, 1969.

 

Grabill, Wilson H. and Lee Jay Cho. "Methodology for the Measurement of
Current Fertility from Population Data on Young Children," Demography.
Vol. II, 1965, pp. 50—68. '

 

Hagget, Peter. Locational Analysis in Human Gepgraphy. St. Martin's Press,
New York, 1966.

Hilton, T. E. Population Atlas of Ghana. Legon, 1960.

Hunter, John M. "Regional Patterns of Population Growth in Ghana 1948-
60," Essays in Geography for Austin Miller. J. B. Whittow and
P. D. Wood (eds), 1965, pp. 272-90.

Johoda, Gustew. "The Social Background of a West African Student Popula—
tion I and II," British Journal of Sociology. Vol. V, pp. 355-69;
V010 V1, ppo 71-9, 19540

King, Leslie J. Statistical Analysis in Geography. Prentice-Hall, Inc.,
New Jersey, 1969.

Kirk, D. "Factors Affecting Moslem Natality," in Bernard Berelson (ed.),
Family Planning and Population Programs. Chicago, 1966, pp. 561ff.

Kuczynski. The Measurement of Pppulation Growth. New York, 1936.

Lee, Everett S. "A Theory of Migration," Demography. Vol. 3, 1966,
pp. 47-57.

 

118

Malthus, Thomas. Essays on the Principle of Population. London, 1789.

 

Metcalfe, G. E. Great Britain and Ghana: Documents of Ghana History
1806—1957. Thomas Nelson (ed.), 1964, pp. 452—468; 487-508.

 

Monkhouse, F. J. and H. R. Willanson. Maps and Diagrams. University
Paper Backs, London, 1951, p. 260.

 

Morris, R. S. "The Ecology of Epidemic Sleeping Sickness: The Signifi-
cance of Location," Bulletin of Entomological Research, Part I and
11, Vol. 43, 1951.

Musham, H. V. "Fertility of Polygamous Marriages," Population Studies,
Vol. 10, Part I, July, 1956, pp. 3-16.

Nabila, John S. An Historical Geography of Central South-Mamprusi.
Thesesis presented to University of Ghana, Legon, for the B.A. Degree.

Omari, T. Peter. "Changing Attitudes of Students in West African Society
Towards Marriage and Family Relationships," British Journal of
Sociology, Vol. XI, pp. 1970210, 1960.

 

Pool, D. I. The Ghana Fertility Survey. Paper presented to the first
African Population Conference, Ibadan, Jan., 1966.

 

Ravenstein, E. G. "The Laws of Migration," Journal of the Royal Statis-
tical Society, Vol. 48, 1885, pp. 167—235; Vol. 52, 1889, pp.
241-305.

 

Raeburn, John R. Report on a Preliminaty Economic Survey of the Northern
Territories of the Gold Coast. Colonial Office, 1950, p. 18.

 

Ryder, Norman B. "Fertility," in Philip M. Hansen and O. D. Duncan (eds.)
The Study of Population: An Inventory and Appraisal. University of
Chicago Press, Chicago, 1959.

Report on the Northern Territories of the Gold Coast. Colonial Office,
London, 1937.

Sessional Papers XXIV, Ropers Relating to the Construction of the Railway
to the North from Kumasi, 1928-1929. Colonial Office.

 

Scott, David. "The Epidemiology of Sleeping Sickness in Ashanti," (in
three parts) Journal of Tropical Medicine and Hygiene. Vol. 60,
1957.

Smith, R. T. Agriculture in South—Mamprusi. Agriculture Department, 1945.

Smillie, K. W. An Introduction to Regression and Correlation. The Ryerson
Press, Toronto, 1966.

Spengler, Joseph J. and Otis Dudley Duncan, (eds.). Population Theory
and Policy. The Free Press, Glencoe, Illinois, 1956.

 

119

Strong, Richard. Diagnosis, Prevention and Treatment of Tropical Diseases.
New York, 1945, p. 164.

Thomas, Edwin H. "Areal Associations Between Population Growth and Selected
Factors in the Chicago Urbanized Area," Economic Geogpaphy, Vol.
36, 1960, pp. 158-1970.

Trewartha, Glen T., "A Case for Population Geography," Annals of the
Association of American Geoggaphers, Vol. 43, No. 2, 1953, pp.
71-79.

 

"A Geography of Population Characteristics," in
Geography of Population. Paul F. Griffin (ed.), Fedron Publisher
1970, pp. 39-640

 

Wrong, Dennis. Population and Society. New York, Random House, 1962.

Zelinsky, Wilbur. A Bibliographic Guide to Population Geogpaphy. Depart-
ment of Geography Research Paper No. 80. University of Chicago Press,
Chicago, Illinois, 1962.

"mummy"!!! mum 111qu "W ITIHITII ES

 

2408