Front Cover
 Table of Contents
 List of Tables
 List of Figures
 Study area characteristics
 Analytical model development
 Empirical findings
 Summary and conclusions

Group Title: Economics report - University of Florida Agricultural Experiment Station ; 89
Title: Agricultural diversification and export earnings
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00027548/00001
 Material Information
Title: Agricultural diversification and export earnings selected African countries
Series Title: Economics report
Physical Description: v, 74 p. : ill. ; 28 cm.
Language: English
Creator: Mathis, Kary, 1936-
Davis, C. G ( Carlton George ), 1936-
Futa, M. T
University of Florida -- Food and Resource Economics Dept
Publisher: Food and Resource Economics Dept., Institute of Food and Agricultural Sciences, University of Florida
Place of Publication: Gainesville
Publication Date: 1977
Subject: Produce trade -- Africa, Sub-Saharan   ( lcsh )
Agriculture -- Economic aspects -- Africa, Sub-Saharan   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Bibliography: p. 70-74.
Statement of Responsibility: W. K. Mathis, C. G. David, M. T. Futa.
General Note: Cover title.
 Record Information
Bibliographic ID: UF00027548
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 000304767
oclc - 04229017
notis - ABT1352
lccn - 78621339

Table of Contents
    Front Cover
        Front Cover
    Table of Contents
        Page i
        Page ii
    List of Tables
        Page iii
        Page iv
    List of Figures
        Page v
        Page 1
        Page 2
        Page 3
    Study area characteristics
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
    Analytical model development
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
    Empirical findings
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    Summary and conclusions
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
Full Text
October 1977

Economics Report 89

Agricultural Diversification and Export

Earnings, Selected African




Dod and Resource Economics Department
agricultural Experiment Station
istitute of Food and Agricultural Sciences
In Cooperation with
inter for African Studies
university of Florida, Gainesville 32611

W. K. Mathis

C. G. Davis

M. T. Futa



This study focused on four African countries: Kenya, Nigeria,
Tanzania and Zaire. The objectives were (a) to identify whether agricul-
tural export earnings fluctuations were determined primarily by quantity
or price variability; (b) to investigate trends and patterns in the agri-
cultural sector; (c) to determine the impact of export crop diversification
on the level and variability of agricultural export earnings; (d) to in-
quire if there was a gap between domestic supply of and demand for food and
to determine if such a gap is due to overinvestment in the export crop
Fluctuations in agricultural export earnings were found primarily
associated with variations in the quantity exported, which were due in
part to actions of government marketing boards. A secondary source of
export quantity variation was variability in rainfall. All four countries
introduced new crops during the study period, but fewer different crops
accounted for shares of agricultural export earnings, so countries ac-
tually became more specialized. An inverse relationship was found between
export crop diversification and the level of agricultural export earnings.
However, it would appear that in the long run, there is likely to be a
positive relationship.
Demand for food was increasing more rapidly than food supply; the
imbalance was not due to overinvestment in the export sector. Low rates
of growth of export earnings implicitly indicated that the export sector
itself lacked resources.

Key words: Sub-Sahara Africa, export earnings instability, export
crop diversification, economic development, export agriculture, develop-
ment policy.


The authors wish to thank the Rockefeller Foundation for Mr. Futa's
support during his time at the University of Florida. The assistance of
Dr. W. W. McPherson is gratefully acknowledged, both as a member of Mr.
Futa's advisory committee and as a reviewer of this manuscript. Drs.
R. D. Emerson and P. J. van Blokland are also due thanks for reviewing
the manuscript. Three lovely and hardworking ladies typed many drafts
and the final copy, and are due much appreciation: Mrs. Patricia Beville,
Ms. Carolyn Williams and Mrs. Mignonne Winfrey. Mrs. Carolyn Dunham
gathered and processed data, which we acknowledge with thanks.


LIST OF TABLES . . . . . iii

LIST OF FIGURES . . . ...... .

INTRODUCTION ... .. . . 1

Objectives . . . . . .
Country Selection . . . . . 2


Common Features . . . . 4
Kenya . . . . . . 9
Nigeria . . . . 11
Tanzania. . . . ... .. 13
Zaire . . . . . . 15


Literature Review . . . . 17
Variable Specification and Description. . ...... 20
Instability Measures . ... ..... 20
Diversification Index. . . . . 23
Model Specifications for Study Objectives . 24
Objective . . . . . 24
Objective 2 . .. . . . 24
Objective 3 . . . 25
Objective 4 . . . . . 26
The General Objective. . . . .27
Data Sources . . . . . ... 28


Price and Quantity Variability . . . .. 30
Patterns and Trends in Export Crop Diversification, 32
Effect of Export Crop Diversification on Levels and
Variations in Agricultural Export Earnings .. 36
Food Supply and Demand Growth Rate Disequilibrium 41
African Choice. . . ....... . ... 47



Effects of Export Crop Diversification on Land
Allocation . . . 48
Kenya . . . . . 48
Nigeria. .. .... . . ..... 50
Tanzania . . . . . 52
Zaire. . . . . 54
The Use of Agricultural Export Earnings for Development
Goals . . ..... . . .. 54


Summary. . . ........... 57
Conclusions. . . 59

APPENDIX . . . . . 61

BIBLIOGRAPHY .. ... . . . 70


Table Page

1 Agricultural trade value as a percent of total trade
value, selected African countries in selected time
periods. . . . . . . 5

2 Agricultural product value as a percent of GDP,
selected African countries in selected time periods. . 5

3 Position of selected African countries according to
two designated selection criteria, 1971-73 . . 6

4 Earnings from major agricultural exports, constant
dollars, 1950-1972 . . . . . .29

5 Variability in agricultural export earnings attributed
to price and quantity changes, 1950-1973 . .. ..31

6 Export crop diversification indices, 1950-73 ...... .33

7 Average annual rates of decline in diversification
indices, selected periods. . . . . .35

8 Average annual growth rates of quantities exported,
major commodities, 1950-1973 . . .... .35

9 Index values of agricultural export earnings, 1950-1972
(1950=100) . . . . . .38

10 Average annual rates of change in agricultural export
earnings and in diversification indices, selected periods. .39

11 Instability indices of agricultural export earnings,
1950-72. . . . .... .. . 42

12 Differences between annual rates of growth of supply
and demand for selected food items, 1965-1973 . ... ..44

13 Land distribution among food and export crops, 1970. .46

14 Export supply function estimates, Kenya . .. ..49

15 Export supply function estimates, Nigeria. . .51


LIST OF TABLES (Continued)

Table Page

16 Export supply function estimates, Tanzania .... 53

17 Export supply function estimates, Zaire . ... 55

18 Average annual values in current dollars of world and
country exports of major commodities, 1950-73 . 62

19 Earnings from major agricultural exports and total
agricultural exports, current dollars, 1950-73 63
20 Values of commodities exported, current dollars, 1950-
73 . . . .. .. 67


Figure .Page

1 Africa, and Sub-Saharan Countries . . .. 3

2 Index values of agricultural export earnings, 1950-1972
(1950=100) ... .. ..... 7

3 Kenya, with major export crop regions . . 10

4 Nigeria, with major export crop regions ... .. 12

5 Tanzania, with major export crop regions. . ... 14

6 Zaire, with major export crop regions . . 16

7 Export crop diversification trends, 1950-72 (1950=100). 34

8 Average price of food staples in northern Nigeria under
alternative simulated policies, 1965-1980 . .. 45


W. K. Mathis, C. G. Davis and M. T. Futa


The study of trade fluctuations has been of much concern in the
post World War II period. Since then a significant proportion of trade
literature has dealt with theoretical and empirical analyses of fluc-
tuations and general instability of export earnings of less developed
countries (LOCs).
In the presence of problems associated with export earnings fluc-
tuations in general, and/or agricultural export earnings fluctuations in
particular, economists have made extensive analyses of the causes and
effects of instability. Unfortunately, a relatively small proportion
of the literature has dealt with policies to correct instability. Agri-
cultural diversification is probably the most widely discussed stabilizing
mechanism. However, the discussions have generally been theoretical and
did not deal with economic impacts of diversification.


The primary objectives of this study are:
(1) Estimate and categorize export crop earnings variations into
price and quantity components as a means of understanding
their relative weights in program planning and policy forma-

W. K. MATHIS and C. G. DAVIS are associate professors of food and
resource economics at the University of Florida. M.-T. FUTA is a
Rockefeller Foundation Fellow in agricultural economics at Oklahoma
State University.

(2) Describe and measure the patterns and trends in export crop
diversification in each of the countries covered in the study.

(3) Measure and describe the influence of export crop diversification
on the level and variability in agricultural export earnings.

(4) Quantify the gap between aggregate food supply and demand and
determine whether or not any existing disequilibrium between
the two components is related to differential levels of invest-
ment between the export and domestic agricultural sectors.
A secondary and more general objective of the study is to identify policy
lines in the agricultural export sector and to evaluate the impact of
such policies on the overall agricultural development of selected countries

Country Selection

This study analyzes determinants and effects of agricultural export
earnings instability in selected Sub-Saharan African countries during
1950-73. Countries were selected according to two criteria: (a) the
degree to which the country is engaged in agricultural trade and (b) the
relative importance of agriculture in its overall economy. The ratio of
agricultural trade to total trade is used to establish the first criterion.
The ratio of the value of total agricultural product to gross domestic
product (GDP) established the second criterion. The size of this ratio
indicates the overall contribution of agriculture to the national economy,
and thus the relative importance of an external disturbance in agriculture
to the economy.
Both these ratios were calculated for thirteen tropical African
countries, from which Kenya, Nigeria, Tanzania, and Zaire were selected
(Figure 1). Selection of the four countries was based on the average
values of both ratios for the period 1971-1973. For both criteria, coun-
tries were grouped into those with (a) large or (b) small ratios. Numerica'
designation of what constitutes a "large" or "small" ratio is admittedly
a normative (and somewhat arbitrary) decision. For the purpose of this
study, it is assumed that a ratio of agricultural trade value (ATV) to
total trade value (TTV) of 50 percent or more is large, whereas any ratio
less than 50 percent is small. A ratio of agricultural product (AP) to
GDP is assumed to be large if it amounts to 33 percent or more while a

FigurZambia MAalaif a S

Figure 1.--Africa, and Sub-Saharan Countries

ratio of less than 33 percent is small. Here, it is assumed for simplicity
that the national economy consists of three sectors: industrial, agri-
cultural and service. Thirty-three percent of GDP would represent an
equal distribution among the sectors.
According to the country selection criteria established above, Sub-
Saharan African countries may be placed into two broad groupings that
reflect their heterogeneity. Countries grouped on the basis of the rela-
tive trade ratio show definite differences. Kenya and Tanzania are among
countries with higher ratios of agricultural trade value to total trade
value, while Zaire has a smaller ratio (Table 1). Nigeria is typical of
those countries experiencing a shift in status from the first to the
second group.
Divisions based on ratios of agricultural product value to GDP show
similar groupings (Table 2). Kenya and Tanzania have economies based mainly
on agriculture, while Zaire is more dependent on non-agricultural sectors.
Nigeria, although exhibiting a relatively high ratio of agricultural pro-
duct value to GDP, has experienced significant changes. Countries grouped
on the basis of the relative trade ratio (Table 1) show remarkable con-
sistency in groupings based on the agriculture product value: GDP ratio
(Table 2). For example, most countries in the high AIV group are in the
high AP group (Table 3). TTV
All four of the countries chosen for study have experienced substan-
tial fluctuations in export earnings. Constant dollar values of agricul-
tural export earnings of the four selected countries increased from 1950
to 1973, but were vulnerable to sharp fluctuations (Figure 2).


Common Features

The four countries selected for study have certain common agricultural
features. All produce several export crops. Kenya and Tanzania share
many of the climatic and ecological conditions typical of East Africa.
Nigeria and Zaire also have similar agro-ecological conditions in certain
regions, even though they are located in different parts of the African

Table l.--Agricultural trade value as a percent of total trade value,
selected African countries in selected time periods

Country 1956-1959 1963-1966 1967-1970 1971-1973

---------------------Percent-----------.--. -
Study countries
Kenya 65 61 57 52
Nigeria 69 66 56 19
Tanzania 76 80 72 71
Zaire 10 3 7 9
Cameroun 75 75 75 68
Ethiopia 95 98 93 90
Gabon 3 2 1 1
Ghana 76 72 79 74
Ivory Coast 72 70 67 68
Malawi 90 91 88 89
Senegal 80 87 73 51
Uganda 85 82 82 90
Zambia 4 2 3 1

Source: F.A.O., Trade Yearbook.

Table 2.--Agricultural product value as a percent of GDP, selected African
countries in selected time periods

Country 1955-1960a 1963-1966 1967-1970 1971-1973
-----------------------Percent--- ---------------
Study countries
Kenya 36 37 33 33
Nigeria 60 59 54 39
Tanzania 50 55 40 39
Zaire 30 21 21 19
Cameroun 40 41 37 36
Ethiopia 57 57 52 54
Gabon -- 17 15
Ghana -- 70 46
Ivory Coast -- 59 33
Malawi -- 51 51 50
Senegal 40 30 28 27
Uganda 60 59 56 53
Zambia -- 9 8 8

aThese years differ
between original data

from those in Table 1

(1956-59) due to differences

Source: U. N., Survey of Economic Conditions in Africa.

Table 3.--Position of selected African countries according to two designated

selection criteria,

AT(a (Lw ATVa AP b AP b
Country (High TTV) (Low (High DP)' (Low 0b

Study countries

Kenya + +
Nigeria + +
Tanzania + +
Zaire + +

Cameroun + +
Ethiopia + +
Gabon + +
Ghana + +
Ivory Coast + +
Malawi + +
Senegal + +
Uganda + +
Zambia + +


= Agricultural Trade Value.
50 percent.

TTV = Total Trade Value

= Agricultural Product.
33 percent.

High T-V means a ratio of 50 percent or more; Low, less than

High GDP

means a ratio of 33 percent or more; Low, less than

GDP = Gross Domestic Product.

Source: Computed from U. N., Survey of Economic Conditions in Africa
Trade Yearbook.

and F. A, 0.,




...-- Nigeria

........... Tanzania

...-... Zaire

I I I I I t 1 I i I t I ft


t t t t l

I I I I I I I I a I
A a a L t a t A I







1955 1960 1965 1970

Figure 2. Index values of agricultural export earnings, 1950-72 (1950=100).

-- --


continent. These common features explain the historical development of pair
oil and natural rubber production in both countries. Coffee, tea and
hard fibers are common to Zaire, Kenya and Tanzania. Similarities between
the agricultural economies of Nigeria and the two eastern countries are
found mainly in institutions introduced in all three countries by the
former British colonial administration.
A common and important institutional legacy of the colonial era is
the marketing board. In fact, agricultural marketing and trade in most
African countries is controlled by these boards. Boards are statutory
bodies established by government action and endowed with legal powers
over the production, marketing and processing of primary agricultural
products (Abbott). However, the objectives, structure, conduct and
performance of such bodies vary from country to country.
Abbott, after extensive analysis of marketing boards, indicates that
these bodies have several objectives. Some of the more important are
sales promotion, research, extension services, raising bargaining power
of agricultural producers in domestic or export markets, setting up needed
marketing and processing facilities, equalizing returns from sales in
different markets or through different outlets, and cushioning the impact
upon producers and consumers of sharply fluctuating internal and external
prices. The last point is of particular interest to this study, since
it deals with problems of agricultural trade instability.
Most African marketing boards are of the export monopoly type. The
Zairean boards are an exception, however, since they are essentially
stabilizing rather than trading institutions (Abbott). In the remaining
three countries, marketing boards handle all problems related to the
marketing of export crops. Kenya and Tanzania have had individual boards
for each export crop. In recent years, Tanzania has been reducing the
number of marketing boards, giving an individual board responsibility for
several crops (Kriesel, et al). In Nigeria, marketing boards are regional
(Helleiner, 1966). Also, two Nigerian export crops, rubber and bananas,
are not regulated at all.
In spite of differences among countries, there are striking similar-
ities in the conduct of African marketing boards. In all four countries,
marketing boards establish annual producer prices, which change little

during a season or from year to year. Where these prices are lower than
world prices the differences accrue to the marketing boards. These sur-
pluses are supposed to be used to maintain producer prices when world
price levels fall below the marketing board price. However, the use of
such surpluses in Nigeria has been criticized by some economists (Nixon,
Eicher, 1971, Johnson, 1969, 1971), and defended by others (Helleiner,
Aspects of this debate will be explored later in discussions of the
empirical findings of this study. This debate, however, is not limited
to the use of trade surplus. It has been extended to the general impact
of marketing boards on market and production mechanisms. There is general
consensus that marketing boards tend to create distortions in factor mar-
kets and result in production disincentives (Johnson, 1968, Helleiner,


Kenya's export crops are produced in three distinct agro-ecological
zones (Figure 3). The first is the eastern coastal plain, a semi-arid
band along the Indian Ocean, where sisal is produced mainly on planta-
tions. The second zone in the interior to the north is characterized
by poor pasture lands and marginal cotton production. The third ecological
zone is a fertile upland region which produces coffee and tea, and most
of Kenya's other export crops. Four crops were selected for this study
because of their relative importance in the Kenyan economy: coffee, tea,
cotton, and sisal, which account for about 80 percent of agricultural
export earnings.
Kenya's major export crops are produced by both plantation and
peasant farming systems. However, the plantation economy remains a
distinguishing feature of Kenyan agriculture. It has only been within
the last two decades that the peasant farming system was given the neces-
sary economic incentives to expand under the auspices of the Swynnerton Plan.

1This plan was introduced in 1954 and supplemented in 1962. The
primary objective was to help the natives of Kenya to expand their pro-
duction of crops such as coffee, pyrethrum, tea, maize and millet. Lands
made available through this plan were divided into two areas--'Scheduled'
areas and 'Non-scheduled' areas. The former was reserved for Europeans
ana the latter for African natives.

Coffee and tea



Figure 3.--Kenya, with major export crop regions

Available lands were designated as 'high-density areas' and were reserved
for subsistence small holders.2 "Low-density" areas were reserved for
larger commercial farming units. These large units market their crops
through marketing boards.


Nigeria, in contrast to Kenya, has large areas of fertile coastal
land. This region can be divided into two subregions: the southeast,
with palm oil and natural rubber production, and the southwestern region
which has maintained its comparative advantage in cocoa production (Figure
4). The second region, stretching from east to west across the center
of the country, is suitable for growing both groundnuts and cotton. The
third zone in the north lacks the climatic advantages of the south but
has been mainly responsible for making Nigeria the world's major exporter
of groundnuts. The four crops selected for study--groundnuts, cocoa,
palm oil, and natural rubber--are produced in all regions and contribute
approximately 85 percent of the total agricultural export earnings of the
Nigeria has areas both of surplus labor and surplus land. It also
has areas, with differing labor-cultivated land ratios and lengths of
fallow periods, between these two extremes (Helleiner, 1966, p. 55).
The coastal region with its high population density and limited land
is considered to be a surplus labor area. The central and northern regions,
on the other hand, with low population density are considered to have
surplus land. Despite the apparent land surplus, these regions are
characterized by permanent and seasonal labor migration to the south
With such differences in regional labor-land ratios, it is easier
to understand why the trend toward individual land proprietorship is
not very significant in the country as a whole (Johnson, 1968). However,

The term "subsistence holders" means farmers consuming more than
50 percent of their farm production.



r-7/7 Palm oil

and rubber

.:.': Groundnuts

Figure 4.--Nigeria, with major export crop regions

in the area of land reform Johnson argues that extended and community
ownership of land is not a serious problem in Nigerian agriculture. He
suggests that the more serious problem is the market distortions caused
by government and marketing board policies (Johnson, 1968). This view
is not generally shared by all economists concerned with Nigerian econo-
mic development. The debate on the role and impact of marketing boards
on Nigerian economic development is still unsettled.


Tanzania, south of Kenya and east of Zaire, has four main regions
of interest (Figure 5). The first is the eastern coast where sisal is
produced. Next is the west-central region, in which several crops are
cultivated, with cotton, coffee, and tobacco the most important. The
north-central region also produces coffee, while the southern zone is
characterized by a system of production which has been referred to as
crop dispersion, rather than crop diversification (Saint-Marc, 1968).
Cash crops grown in this system include sesame, tobacco,.cotton, tea,
sunflower, castor seed and cashews. Coffee, tea, cotton and sisal--the
four study crops--account for approximately 75 percent of the agricul-
tural export earnings of Tanzania.
Export crop production is largely concentrated in the plantation
economy. However, as in the case of Kenya, emphasis is now being placed
on the development of small farms. The Tanzanian government as well as
the Tanzanian national political party maintain a strong interest in the
socialization of the rural society.5 Governmental participation has also

3Dispersion is defined as the existence of several crops in a given
area with a very low density.
4Cashews are important Tanzanian exports but lack of data prevented
inclusion in this study.
Rural development policy is expressed as a program of rural social-
ism. It is designed to lessen income inequalities among farmers by giv-
ing them a community mode of production by which the farm unit is jointly
owned by the extended family or a group of people who agree to work to-

Coffee and tea
LU'I I Sisal
XX Cotton

Figure 5.--Tanzania, with major export crop regions

been extended to include the marketing system. This system operates
through a complex set of marketing regulations administered through
marketing boards. The marketing boards' structure and operation have
been reviewed and subjected to a number of changes over the years,
largely as a result of criticism of African governments' use of agri-
cultural surpluses. However, in the case of Tanzania, marketing board
surpluses have been used for the economic betterment of the farming sector
(Kriesel, et al). In fact, Kriesel found that there have been no direct
transfers of such surpluses or savings to governments in contrast to the
case in some West African countries.
Thus, in spite of some uneasiness about government intervention in
the agricultural system, there is optimism that agricultural surpluses
will continue to be invested in the agricultural sector instead of being
siphoned off to other sectors. However, government policy encourages
community ownership as the means of production. With a high population
growth rate of 2.7 per year, Tanzania has a serious shortage of cultivable


Zaire, the third largest African country after Algeria and the Sudan,
has the most diversified climate of any country on the continent. The
varied ecological conditions are suitable for many different crops, but
only a few are grown commercially for export. Coffee, tea, palm oil, and
natural rubber represent 85 percent of all the agricultural export earn-
ings of the country (Figure 6).
These crops are produced mainly on plantations, but there'are some
small peasant farm units. In the 1950's the colonial agricultural admin-
istration began a program known as paysannats indigenes in an effort to
transform traditional farming patterns. Innovations implemented in these
programs were noted in development literature (Johnson, 1968; McPherson
and Johnston, 1968). However, the system collapsed with the emergence
of national independence. The collapse of the paysannats system was
accompanied by a decline in the relative importance of agriculture in the
economy. Massive migration has been occurring from rural to urban areas
(Mabala). A prevailing pattern is that of disinvestment in the agricultural
sector (Peemans).
In addition, agricultural institutions in the country have not been

*t -'i~. *" i
*# .4 t j ,
*~.> > ~.s

x Y Xx
x XXx
xX \XXm

r:i Natural

|(| Palm oil
z Coffee a

Figure 6.-- Zaire, with major export crop regions.


nd tea

effective in rural improvement programs. However, the Zairean government
has begun to provide credit facilities, revise land policies, and improve
the rural infrastructure.


Literature Review

In the wake of post World War II interest in trade fluctuations,
several empirical analyses were undertaken. A 1952 United Nations study
reported an analysis of price and quantity movements in LDCs. The study
focused on: (a) year-to-year price and quantity fluctuations and (b)
long-term price and quantity fluctuations and cyclical swings. One of
the major findings was that export earnings fluctuations tended to be
higher than those of prices and quantities taken individually, due to
the interaction of prices and quantities. The study also reported that
prices accounted for two-fifths of the fluctuations-while volume accounted
for the remaining variability. Using United States export data, Mintz
concurred with the United Nations findings regarding the relative impor-
tance of export quantity as a determinant of variability in export earnings.
In a 1958 article, Nurkse formulated specific policies on the basis
of the earlier United Nations report. He supported the United Nations
proposal calling for international funds and buffer stocks as a means of
stabilizing export earnings. However, he made a strong case for balanced
growth in the LDCs as a means of fostering industrialization and less
reliance on primary products.
Coppock, in 1962 and later in 1966, analyzed Middle Eastern foreign
trade patterns. In both studies, Coppock developed indices of instability
and related them functionally to export prices, export quantities and
market shares of individual commodities and countries.
A study by DeVries concluded that such factors as trade position,
price inflation and resource allocation determined export growth rate and
performance in LDCs. DeVries found a positive correlation between the
performance of major and minor exports and growth in the value of agricul-
tural product. He opposed diversified industrialization policies as

inefficient in helping LDCs reach economic levels of production and a
competitive ceiling.
Studies by Maizels have made significant contributions to the under-
standing of trade problems in LDCs. Maizels argued against overspeciali-
zation in particular export commodities. He saw export diversification
as not only an appropriate mechanism for facilitating structural change,
but as probably the most important from a long-term viewpoint. Maizels
further argued that accurate assessment of world demand trends for export
commodities is a vital first step in assisting LDCs to capture the econo-
mic gains from high demand exports.
Balassa's work had significance for the agricultural earnings insta-
bility question in that it estimated demand trends for temperate zone
foods, competing tropical foods and agricultural raw materials. However,
an increase in world demand might not justify a policy of export crop
diversification and shifts in resource allocation. Some economists have
proposed reorienting policies towards food crop diversification as a means
of reducing high food import propensity (Flores).
All of the studies discussed so far, except that by the United Nations
dealt with trade problems in general or with the question of stability,
without referring to specific stabilization policies. Massel entered these
gaps by reviewing different policy alternatives. Buffer stocks and multi-
lateral contracts proposed by the United Nations study and supported by
Nurkse were analyzed. Massel found that buffer stocks provided certain
welfare advantages to producers by minimizing the expected value of change
of producer prices, and to consumers by providing gains in consumer surplus
(1969). However, the cost of implementing a buffer stock policy was higher
than other alternatives (Massel, 1970). Earlier, Massel concluded that
neither export earnings instability nor the disutility arising therefrom
are likely to be eliminated by simple policies such as diversification of
exports (1964). Since variations are independent among commodities, their
additivity may actually worsen the variability. This argument runs counter
to Jorberg's findings that diversification is capable of inducing stability
in the secular trend of export earnings.
Parikh used an econometric model of the world coffee economy to predict
production, consumption and outcomes of alternative policies. A similar

model was later used by Edwards and Parikh to identify policies that would
minimize the fluctuations of agricultural export earnings. They found
that, given the necessary resources, an international buffer stock policy
could substantially reduce short-run fluctuations. A quota policy was
judged to be more successful, as it had fewer of the economic difficulties
found in the buffer policy. Edwards and Parikh suggested, however, that
quota policies would probably be much more difficult to enforce, due to
political considerations.
With such difficulties evident in international policies, interest
should turn to domestic policies such as tax structure and export crop
diversification. However, little has been done in this regard in LDCs,
and the concept of diversification has been primarily associated with
the process of industrialization. Only a limited number of studies have
analyzed export crop diversification as such. One of the few international
studies dealing with this aspect is a 1967 report by the Committee for
Economic Development (CED). This particular study concluded that export
earnings fluctuations are largely the result of a combination of varia-
tions in crop output and dependency on one or two products.
Policies and programs have been formulated on the basis of many of
the above analyses in an attempt to facilitate greater international price
stability and economic growth (U.N., 1952). The major operating mechanisms
of these stabilizing programs were export quotas, buffer stocks, and multi-
lateral contracts. The General Agreements on Tariffs and Trade (GATT) has
served as the primary operational vehicle for these policies and programs.
In spite of GATT's efforts, the LDCs still exhibit instability in export
earnings, particularly for agricultural exports.
The problems of agricultural export earnings instability are likely
to be more serious in those countries that derive a sizable proportion of
their Gross Domestic Products (GDP) from agricultural exports. Although
it has not been conclusively determined that instability hampers economic
development (Lim), it has been established that fluctuations in agricultural
export earnings have a multiplier effect on domestic incomes. Fluctuations
in earnings accentuate inflationary and deflationary movements in the
national economy, and may also discourage investment in the agricultural
sector and thus impede the growth process.
Instability of agricultural export earnings can also inhibit government

programs aimed at meeting social welfare goals. These variations may force
a country to borrow in order to meet its social welfare objectives. Where
loan repayment obligations accumulate over time, additional stress on the
nation's economy may result. Thus social efficiency may be reduced if
the agricultural export sector generates a particular system of resource
misallocation (Beckford).

Variable Specification and Description

Instability Measures

Attempts to quantify the variability of export earnings have been made
by several authors. Instability is used here to mean the deviations of
the constant dollar values of agricultural export earnings around the trend.
The United Nations study (1952), referred to earlier, used an average of
year-to-year percentage changes as a measure of instability. The mathe-
matical formula of that index was:

N (t+ Z x 100 ()
IUN ____t


IUN = United Nations instability index,

Zt = real value of agricultural export earnings in period t, and
N : number of years for which the instability is computed.
Kingston criticized this procedure by arguing that it tended to
exaggerate the importance of the instability existing in the time series
(Kingston, p. 20). He wrote that:
One obvious deficiency of this approach is that a
steady increase of a constant percentage per annum
would be interpreted as an unstable movement, when
in fact there had been no instability, in a conven-
tional economic sense, at all. Rather, there would
have existed a stable percentage of growth.
Coppock, who had already criticized the United Nations index with
the same arguments as Kingston, formulated an index with a logarithmic

variance measure, to account for a constant percentage growth rate. The
Coppock index may be written as:

ICp = antilog z (log Z.t -M) (2)
...... Zt

ICP Coppock instability index,
M = the arithmetic mean of logarithm differences, and
Z,N = as defined in equation (1).
While the Coppock index may represent an improvement over equation
(1), it does have some weaknesses. First, the log-variance form interprets
small deviations from a low base as highly unstable. Second, instability
coefficients derived from percentage changes are difficult to interpret
Another instability index was proposed by Massel, who used a least-
squares regression model with time series data. Massel's model is de-
scribed mathematically as:

N 2
IM = (Zt Zt) (3)
t=l N

IMS = Massel index of instability,
Z' = least squares estimate of Z, Zt f(t),
Z = trend mean value of agricultural export earnings,
Z,N = as defined in (1) and (2).
This procedure does not insure that the least squares model used to
estimate Z4 is the correct model, but it is an improvement over (1) and
(2). Staller, as cited by Kingston, applied this index with the slight
modification of regressing the logarithmic values of Zt on time.
An interesting aspect of the Massel index is that it may be trans-
formed to accommodate analysis of times series data. Previous indices
were limited to measuring total instability over a given N years. The
transformation of Massel's index in a time series basis measures the
instability of period-to-period changes. It can be stated as follows:

I' = u u (4)
MS t+1 t
(Zt+l, t)

IMS = transformed Massel annual index of instability,
ut+l = residual error corresponding to Zt+l observation,

ut = residual error corresponding to Zt observation.

Since this transformed version of Massel's index applies to a time
series model, it was used in the present study. However, some modificatior
were needed in the denominator of equation (4). Massel implied that the
point from which deviations occur is itself a changing or a dynamic point.
But, the weakness of the assumption is that at each point in time, there
exists a given maximum value which is considered as a stationary value.
This is misleading, since it is known that there are some conditions to
be met for a point to be considered as a stationary value (Gandolfo).
Thus, in order to avoid errors that the second Massel index may generate,
equation (5) was developed. Such a relation takes the form:

It t+l t (5)

It = corrected index of instability,
ut+l = residual as defined in equation (4), but obtained with the
exponential trend in equation (10).
ut = residual as defined in (4), but derived with the exponential
trend in equation (10).
Z = average trend value from equation 10, considered as a fixed
stationary value, from which deviations take place.
It should be noted that all these indices measure the variability
about a given stable point. They do not, however, possess the properties
of normal variances. Thus, these constructs have a basic shortcoming that
prevents their application to analysis of the different effects of price
and quantity variation on agricultural export earnings instability. To
separate the price and quantity components of export earnings variation,
normal variance properties must be included. One of the properties of
the normal variance of a given variable is that such a variance may be

apportioned into two or more components.

Diversification Index

Diversification refers to the production of several crops in a given
geographical area or production unit. But, if only the numbers of products
are considered, Finke and Swanson argue that diversification can be thought
of as a "richness" concept meaning 'how many' crops a given farm unit grows.
However, emphasis may also be placed on the relative importance of a given
crop in a given area or farm unit. In this case, diversification could
mean that all crops in the area are equally or evenly distributed. Defined
in this way, diversification is the degree of 'evenness', as determined
by the proportions of resources allocated to each crop planted.
The Committee for Economic Development (CED) proposed that agricultural
diversification be thought of as a dynamic process, with new crops intro-
duced into the system over time. The CED defined the diversification
index as the ratio of the growth rate of agricultural export earnings to
the growth rate of export earnings from traditional export crops. This
definition, although interesting, has some shortcomings when used in re-
gression models where export earnings are dependent variables. Thus the
index used in this study is that developed by Finke and Swanson. It can
be formulated as:
Dt = -n E li log (li) (6)
Dt = diversification index,
n = number of crops, and
li = proportion of land resource allocated to crop i.
However, because crop data in this study are based on harvested area
rather than on actual planted area, and because the study is primarily
focused on the export sector, it was felt that the substitution of the
proportion of total agricultural exports represented by each crop for the
proportion of land allocated was more appropriate. Therefore, equation
(6) may be adjusted to:
Dt = -n z qilog (qi) (7)
qi = share of crop i in total value of agricultural exports.

Objective 1

Fluctuations in agricultural export earnings are decomposed into
three components: price variations, quantity variations and variations
resulting from the interaction between prices and quantities. An
approximation of the variance of the function, Zi = PiQi, following the
procedure used by Motha and others, is:

I2 -2 (2 -p2 o2 pa a (8)
zi Pi P iq pi aqi

Oa = agricultural export earnings variability of crop i,

Qi = average quantity of crop i exported, in thousands of
metric tons,
P. = average export price of crop i exported, in U.S. dollars,
a = price variability of crop i,

a2 = quantity variability of crop i, and

p = correlation coefficient between quantity Qi and its
price P..

The first term on the right hand side represents price variability,
the second term quantity variability and the third is the interaction tern
Higher order terms in the series expansion are not included. Initial
investigation showed negligible covariance between price and quantity, so
those covariance terms are not included in further discussions.

Objective 2

In order to identify and quantify the trends and patterns of export
crop diversification, the exponential trend procedure was adopted because
of the good fit exhibited in preliminary trials. The symbolic represen-
tation of the model is:
Dt = aeT (9)

Dt = export crop diversification as defined in equation (7),
a = scale of export crop diversification,
S= growth rate of export crop diversification, and
T = time trend.
The parameters a and o are calculated by the trend procedure.

Objective 3

The third objective, evaluating the impact of export crop diversification
on the level of agricultural export earnings, measures trends in agricultural
export earnings obtained by the exponential trend procedure in objective 2.
It takes the form:

Zt = ae fT (10)
where all variables are as previously defined. The impact of diversification
is measured with a simple linear regression model and a quadratic regression
model. The simple model is used to identify the nature of the relation-
ship between two specific variables, while the quadratic model was adopted
as a means of simulating the impact of intensified diversification. The
basic regression form may be written as follows:
Zt = f(Dt, ut) (11)
where variables are as defined above.
Expansion of the two models gives two separate equations:
Zt = bo + bDt + Ut (12)
Zt = bo + blDt + b2Dt + Ut (13)
A second aspect of this objective deals with evaluating the effect
of export crop diversification on the level of instability. This is
accomplished by applying equations (12) and (13) to the instability index,
i.e., by regressing the instability indices upon the diversification vari-
ables. The resulting equations are functionally similar to equations (12)
and (13). They have the following form:

It = ao + aiDt + ut (14)
It = ao + alDt + a2D + ut (15)
Equations (12) and (14) are simplistic and are only described for
theoretical considerations. For interpretation purposes, emphasis is

placed on equations (13) and (15) which correspond to the corrected in-
stability index given as equation (5).

Objective 4

The rationale behind the fourth objective was to determine if the
export crop sector was developing at the expense of the domestic food
sector. A simple equilibrium equates the growth rate of food supply to
the growth rate of food demand. It is assumed that if there is disequi-
librium on the supply side, there are factors inhibiting the growth of
the domestic food sector. One such factor could be a significant dif-
ferential between the levels of investment in the export crop and in the
domestic food crop sectors. Investment is used here in its broad sense
and covers such things as land and capital allocation, marketing insti-
tutions, research infrastructure, labor and other relevant inputs. Be-
cause of data limitations, only policies affecting land allocation be-
tween the two sectors were examined. It is likely, however, that a high
correlation exists between the quantity of land and the quantity of labor
used in a particular sector.
The relation used here considers the supply side as consisting of
domestic food production and food imports. The demand side is formulated
as proposed by several economists (Okhawa, Johnston and Mellor, 1961).
The detailed equilibrium equation is:
Sid + Sif = H + niG (16)

Sid = growth rate of domestic food production of food item i,
Sif = growth rate of food imports of item i,

H = population growth rate,
n. = income elasticity of demand for food item i, and
G = real income per capital.

Rice, maize, beans and sugar were selected for study, since these
commodities are basic food items and estimates of income elasticities
are published (FAO, 1967).6 While it would have been appropriate to

6There are some inherent shortcomings of these estimates, since they
are mainly projections and not actual estimates for specific time periods.

include several staple African food crops, such as cassava and yams, there
are no data on income elasticities for these items.

The General Objective

In order to identify the determinants of some of the policies pre-
vailing in the African export crop economy, it was felt that understanding
the impact of individual crops on the level of agricultural export earn-
ings was an important first step in explaining how land resources are
allocated. Such an impact may be measured through a multiple regression
analysis model of the type:

t f(Qit, Dt' uzt) (17)

Zt = level of agricultural export earnings as defined previously,
Qit = quantity of crop i, in thousands of metric tons exported
in period t,
Dt = diversification as defined in (7), and
uzt = error term.

However, Qit is itself an endogeneous variable which depends on such
predetermined variables as export prices, harvested area of export crops,
weather and time trend. The appropriate model is a system of simultaneous
equations, where the system is identified, and a two-stage least squares
procedure is used. The system has the following form:
t = f(it, Dt, Uzt) (17)

Qit = fit-l' Pot-,l Ait, Aot, RFt T, uqt) (18)

Pit-l = export price of crop i in period t-l, in U.S. dollars,
Pot-_ = average export price of all other export crops in period
t-l, in U.S. dollars,
Ait = harvested area of crop i in period t, in thousands of
Aot = harvested area of all other export crops in period t,
RFt = rainfall index in period t,

T = time trend,
UztUqt = error terms.
While price variables in the system above would give information
regarding price responsiveness, Ait and Aot indicate relationships
between changes in the area of a given crop, relative to changes in
areas of others. Such information is very important for explaining
the patterns of allocation of cropland between the crops selected for
study. It also provides some insights into the relative degree of
scarcity or abundance of land resources.
Weather, particularly rainfall, frequently causes quantity varia-
bility in export earnings, so a rainfall index was incorporated. How-
ever, this index can only serve as a proxy for weather effects, and
might not show a direct relationship with quantity exported. For example,
domestic storage and processing of some export commodities means that the
quantity exported in any given year is not equal to the total quantity
produced. The quantity produced would be responsive to weather varia-
bility, while the quantity exported would not. Irrigation and flood
control schemes could also modify the effects of rainfall on export

Data Sources

The study used secondary data from F.A.O. Trade and Production
Yearbooks, United Nations Surveys of Economic Conditions in Africa,
International Monetary Fund Surveys of African Economies and country
statistical abstracts and reports prepared by different missions. Export
values in U.S. dollars (Table 4) were deflated for all calculations. The
deflation procedure used was:
current dollar exports (y) = implicit deflator (b)
constant dollar exports (x) 100
so that
constant dollar exports (x) = y (100) .

The implicit deflator used for each country was the mean value of deflators
for each year calculated in the United Nations' Survey of Economic Con-
ditions in Africa.

Table 4.--

Earnings from major agricultural exports,a constant dollars,

Year Kenya Nigeria Tanzania Zaire

-------------------M-llion U. S. dollars----------------

1950 9.1 101.0 40.0 105.0
1951 7.5 76.8 37.4 80.0
1952 9.0 70.8 40.2 100.0
1953 11.0 48.0 41.8 102.0
1954 7.5 54.0 33.0 102.0
1955 8.0 66.0 42.0 118.0
1956 9.0 72.0 42.0 140.0
1957 10.0 90.0 42.0 148.0
1958 10.0 96.0 42.0 152.0
1959 9.5 90.0 46.2 164.0
1960 10.4 96.0 45.7 166.0
1961 15.0 96.0 46.4 160.0
1962 13.5 108.0 46.4 142.0
1963 14.0 120.0 55.0 142.0
1964 16.0 150.0 56.2 144.0
1965 21.0 144.0 52.8 142.0
1966 20.0 162.0 60.5 140.0
1967 18.0 189.0 70.4 126.0
1968 22.0 150.0 56.1 120.0
1969 20.0 180.0 68.2 124.0
1970 22.5 189.0 70.4 112.0
1971 21.0 150.0 77.0 108.0
1972 20.0 149.0 77.0 84.0

aMajor exports:
Kenya, Tanzania Coffee, tea, cotton and sisal.


- Groundnuts, cocoa, palm oil and rubber.
- Coffee, tea, palm oil and rubber.

Source: Computed from F. A. 0., Production Yearbook and Trade Yearbook,
selected years.


In interpreting the empirical findings, a specific conceptual frame-
work was utilized. That framework was one that maintains that export
price stability depends mainly on two factors: (a) the fluctuations in
demand for or supply of export commodities, and (b) the efficiency of
existing stabilization schemes (Rowe). Evidence suggests that exports
with relatively stable prices (compared to quantity variations) are those
that continue to be supported by international commodity agreements.
Tea, although protected up to 1955 by the International Tea Agreement,
has since been subjected to free trade. This change in status largely
resulted from the assumption and belief that its final demand and pro-
duction were very inelastic. Sisal, on the other hand, has never been
individually chartered by an international commodity control scheme. It
is, however, a part of an international Conference of Hard Fibers Study
Group sponsored by F.A.O. The movement of the price of sisal after the
Second World War has been characterized by substantial ups and downs.
Sisal price was low during the interwar period, rose in the first half
of the 1950's, then fell until the mid-1960's (Rowe, Kriesel et al).
The objectives of this study did not allow further investigation of the
effects of international commodity agreements on price stability and
export earnings of African countries. This subject might, however, be
a fruitful one for further research.

Price and Quantity Variability

From the estimating procedure given in equation (8), it was found
that of the three study countries for which coffee is a common export
crop, two countries (Kenya and Tanzania) exhibited a relatively higher
percentage of quantity variability compared to price variability (Table 5).
For Kenya, Tanzania and Zaire, three countries for which tea is a common
export crop, results indicate that in two of these three countries (Kenya
and Zaire) price affects variability more than does quantity. Price
variability is proportionally more important for sisal, a crop common to
Kenya and Tanzania. Other common crops such as cotton, palm oil and rubber,
show that quantity variability is the most important component of the

overall variability (Table 5). Thus of six common crops among the countries,
four crops (coffee, cotton, palm oil and rubber) were found to be charac-
terized by quantity variability, whereas two (tea and sisal) were subjected
to higher price variability.
For these four countries, cocoa and groundnuts are major exports only
in Nigeria. These commodities are affected more by quantity variability
than by price fluctuations. In the aggregate, the evidence suggests that
variability in export earnings is largely associated with the quantity
rather than the price component.

Table 5.--Variability in agricultural export earnings attributed to
price and quantity changes, 1950-73

Kenya Nigeria Tanzania Zaire
Variability attributed to
Crop Price Quantity Price Quantity Price Quantity Price Quantity

----------------------------- Percent------- ---------

Coffee 14 86 35 65 66 34
Tea 55 45 44 56 60 40
Cotton 18 82 4 96
Sisal 73 27 81 1.9
nuts 12 88
Cocoa 34 64
Palm Oil 13 87 35 65
Rubber 14 86 25 75

These results are consistent with findings in two other studies (United
Nations, 1952; Mintz).
In addition to the price and quantity variability calculated, an
interaction component was computed. The meaning of this component has
been subjected to many interpretations (Motha, et al). In this study,
it is simply interpreted as the variability resulting from interaction
between price and quantity variations. Its impact on the total varia-
bility stems from the fact that, if the correlation coefficient between
price and quantity variations is either positive or negative, the inter-
action component variability will be, respectively, either positive or
negative. In the former case, earnings variability will be greater than

the sum of both price and quantity variations. In the latter case, it
will reduce the total variability and contribute to the stabilization
process. In this study, the interaction component, whether positive or
negative, was so small that its effect on total variability was negli-
One reason for the small size of this interaction component for each
commodity is the small share in total world exports represented by each
country's exports, in most cases. Only Nigerian exports of cocoa and
palm oil, sisal shipments from Tanzania and palm oil exports from Zaire
accounted for more than ten percent of the value of world exports in each
of those commodities (Table 18).

Patterns and Trends in Export Crop Diversification

While all four study countries may be considered as highly diversi-
fied in terms of the number of export crops grown, the results of this
study indicate that when diversification is defined as given in equations
(7) and (9), all four countries exhibited a persistent and declining
trend in the level of diversification over the study period (Table 6 and
Figure 7). Kenya and Tanzania experienced similar rates of decline of
about one percent per year over the 1950-1973 period. The rates of de-
cline in the level of export crop diversification were also similar for
Zaire and Nigeria, at over three percent annually (Table 7). When the
data series were divided into two time periods, significant differences
between countries were found in the annual rates of decline for the
periods 1950 to 1960 and 1960 to 1973.
The first period, 1950-1960, was one of colonialism, in which policy
emphasis was placed on specialization in agriculture. The second period
was one of political independence and nationalism, with policy orientation
towards the integration of the agricultural export sector into the national
economy. From 1950 to 1960, the export crop sectors of all four countries
were becoming less diversified, but at a decreasing rate. In the later
period, the rate of decline in diversification slowed markedly in Kenya,
Tanzania, and Zaire, but increased in Nigeria. These trends demonstrate
that the patterns of export crop development were determined by a crop
specialization process.
It will be recalled from the section on study area characteristics

Table 6.--Export crop diversification indices, 1950-73

Year Kenya Nigeria Tanzania Zaire






Source: Calculated from F.A.O. Trade Yearbook.

- -- -I



.......... Tanzania


o I ...- ... ...

1 .

.....'... \
- I
Z 100

o 50

I i I It II t
1950 1955 1960 1965 1970

Figure 7.--Export crop diversification trends, 1950-72 (1950=100)


Table 7.--Average annual rates of decline in diversification indices,
selected periods

Country 1950-1973 1950-1960 1960-1973

-------------------- Percent---------------------

Kenya 0.87 1.71 0.19
Nigeria 3.39 2.42 5.73
Tanzania 1.19 1.82 0.82
Zaire 3.26 5.67 3.23

that export crops are distributed in specialized geographical regions.
Furthermore, distributions occur in such a way that competition between
major export crops for land is not great. Thus, the tendency for a coun-
try to concentrate on one or two export crops may be explained in terms
of the persistent growth in the production of certain export crops in one
or two regions, relative to other regions. In the case of Kenya, the
quantities of coffee and tea exports increased more than those of cotton
and sisal since 1950 (Table 8). In terms of regions, it may be argued
that the northern and coastal zones were growing more slowly than the
southwestern region where coffee and tea are produced (Figure 3).

Table 8.--Average annual growth rates of quantities exported, major com-
modities, 1950-73

Crops Kenya Nigeria Tanzania Zaire

------------------Percent per year----------------
Coffee 1.89 -- .86 1.85
Tea 2.49 -- 6.06 2.27
Cotton .33 -- 5.18 --
Sisal .18 .50 --
Groundnuts .30 -
Cocoa -- .68 --
Palm Oil -- -.20 -- .68
Rubber 2.53 .16

Source: Computed from F.A.O., Trade Yearbook, selected years.

Cocoa and rubber led export quantity growth in Nigeria (Table 8).
Cocoa exports, with a rate of growth of .68 percent, were lower than
those from rubber at 2.53 percent. Groundnuts was third with an overall
growth rate of .30 percent, while palm oil exports declined at -.20 per-
cent per year. The high growth rate exhibited by rubber exports was not
steady, but characterized by sporadic increases in quantities exported.
It is unlikely that such a high rate of increase will be maintained,
since more than one half of the area planted to rubber was expected to
be out of tapping after 1974 (Tims). While rubber has been a leading
export, it is felt that cocoa and groundnuts are the major export crops
in which Nigeria has specialized. The southwestern and northern regions
(Figure 4) have steadily increased their exports of cocoa and groundnuts.
In Tanzania, growth rates for tea and cotton were 6.06 and 5.80 per-
cent per year respectively, far greater than those for coffee and sisal
(Table 8). Export growth rates of coffee and tea for Zaire surpassed
those of palm oil and rubber (Table 8). Coffee and tea are mainly pro-
duced in the eastern part of the country, while palm oil and rubber are
mainly grown in the western part (Figure 6).

Effect of Export Crop Diversification on
Levels and Variations in Agricultural Export Earnings

In general, it appears that export crop diversification is working
against high levels of agricultural export earnings. The magnitude and
significance of the regression coefficients from equation (13) for each
country are shown below:
Zt = -3123* -73.51 D* + 8491.18 D2*
(1452) (32.1) (3840)
R2 = .619; DW = 1.679

Zt= -4829* -12.42 Dt + 1349 D2*
(1871.7) (4.90) (512.92)
R2 = .48; DW = 2.33
Zt = -101307** -239.96 Dt* + 27226 D **
(5388) (124.33) (14329)

Tanzania:' continued.

R2= .51; DW = 2.60

S= -2149 -17.23 Dt + 818 D2
(2653) (20.3) (941.3)

R2 = .07; DW = .694

Zt = values of agricultural export earnings as defined in
equation (13),
Dt = diversification index as defined in equation (13),
= b-coefficient significant at the 10 percent level,
** = b-coefficient significant at the 5 percent level,
( ) = values in parenthesis are standard errors.
It appears that, while export crop diversification (Dt) tends to
reduce the level of agricultural export earnings, the long term intensi-
fication of the diversification process tends to shift the earnings to
higher levels. This is the interpretation of the results with the Dt
squared variable defined as intensified diversification. The coefficients
are all significant, except for Zaire; but even there the signs are con-
sistent with those of other countries.
Comparing these findings from equation (13) with those from equations
(7) and (9) yields important conclusions. Export crop production in all
four countries became more specialized between 1950 and 1973 (Table 7).
Since equation (13) shows an inverse relationship between the level of
diversification and export earnings, it would be expected that agricul-
tural export earnings of the four countries increased.
Agricultural export earnings from all four countries grew over the
period (Table 9 and Figure 2). Kenyan earnings grew at higher rates than
those of Nigeria and Tanzania, and all three countries showed much higher
growth rates than did Zaire. The relatively higher rate of growth in
earnings in Kenya was associated with a smaller rate of decline in the
level of diversification, compared with Nigeria and Zaire (Table 10).
Agricultural export earnings grew more rapidly after 1960 in Kenya,
Nigeria, and Tanzania than prior to that year. However, the rate of
earnings growth in Zaire of over 7 percent annually showed a marked

Table 9.--Index values of agricultural export earnings, 1950-72 (1950=100)

Year Kenya Nigeria Tanzania Zaire






Source: Calculated from Table 4.

Table 10.--Average annual rates of change in agricultural export earnings
and in diversification indices, and average instability indices,
selected periods

Country and Unit 1950-1972 1950-1960 1960-1972

earnings percent 6.73 3.63 6.82
cation -0.87 -1.71 -0.19
instability index .35 .33 .39
earnings percent 4.38 1.34 6.09
cation -3.39 -2.42 -5.73
instability index .44 .33 .50
earnings percent 4.13 1.76 5.19
cation "-1.19 -1.82 -0.82
instability index .35 .38 .34
earnings percent 0.30 7.19 -3.17
cation -3.26 -5.67 -3.23
instability index .44 .36 .51

decline after 1960.
Apparently, governments of African countries recognized the inverse
relationship between export crop diversification and the level of export
earnings, and pursued policies of export crop specialization as a means
of maximizing earnings. The direct relationship between intensified expor
crop diversification and earnings shown in equation (13) does, however,
support Jorberg's argument that diversification could be helpful for long-
term growth and development. While diversification reduces the level of
export earnings in the short run, the long-run effects on earnings are
The second aspect of the third objective of the study was concerned
with the effect of export crop diversification on the variation or in-
stability of earnings. On this aspect, the results from equation (15)
are neither conclusive nor consistent, as shown below.
It= 1.68 + 3.42 Dt 1.41 Dt
(7.60) (12.39) (-.283)
R2 = .006; DW = 2.41
I = 5.75 9.37 D 4.08 D
(4.79) (18.07) (3.37)
R2 = .104 DW = 1.95
I = -27.53* + 45.77 D -18.68 02*
(19.02) (14.62) (15.89) t
R2 = .398 DW 1.404
It = 21.53* 48.72 D + 25.19 D2*
(13.20) (36.63) (21.34) t
R2 = .117; DW = 2.17
It = instability index as defined in equation (15),
Dt = diversification as defined previously,
= b-coefficient significant at the 10 percent level, and
( )= values in parentheses are standard errors.

All coefficients in equations for Tanzania and Zaire are significant.
This may indicate that export crop diversification is positively related
with earnings instability in Tanzania, and inversely related to Zaire.
However, R2 values are low for all equations, and all coefficients are
statistically insignificant for Nigeria and Kenya. No conclusive state-
ments can be made. Findings for Tanzania and Zaire may indicate, though,
that diversification might be an appropriate policy for one country,
but inappropriate for another.
Differences between countries could be due to the nature and
characteristics of export crops produced in each. Price elasticities
for each of these crops differ greatly, so that identical changes in
quantities exported could result in very different changes in their
respective revenues and subsequent differences in instability values.
Differences between instability index values for Tanzania and Zaire are
largely due to sisal. Other major determinants are cotton in Tanzania
and palm oil and natural rubber in Zaire, since tea and coffee are
common to both. These observations are supported by the fact that
average instability index values for Kenya and Tanzania, which grow
similar crops, are similar at a .35 (Table 11). Nigeria and Zaire, while
not having all four crops in common, both had average instability index
values of .44 (Table 11).

Food Supply and Demand Growth Rate Disequilibrium

The contention (Fitzgerald, et al) that the export crop sector is
competitive with the domestic food sector for national resources led to
an investigation of the rates of growth in supply and demand of certain
food items. The equilibrium relationship stated in equation (16) was
used with elasticities for food estimated by FAO for 1965-1973 (F.A.O.,
1967). Although these data appear to overstate the growth rates of food
demand, they nevertheless provide some insights and indicate the need for
further investigation.
All four countries exhibited significant disequilibrium between
supply and demand for the selected food crops (Table 12). Nigeria, how-
ever, was self-sufficient in major staple carbohydrate foods such as
yams, cassava, guinea corn, rice and millet (Johnson, et al, 1969).

Tablell.--Instability indices of agricultural export earnings, 1950-72

Year Kenya Nigeria Tanzania Zaire






Average .35 .3 .44-

.35 .44

Average .35

Three of the selected food items (maize, beans, and sugar) belong to what
CSNRD classified as important import substitution crops and nutritionally
superior foods. Sugar is considered an import substitute while maize
and beans are included among nutritionally superior foods (Johnson, et
al,,1969), It seems reasonable to conclude that Nigeria was self-suf-
ficient in staple carbohydrate foods, but experienced shortages in import
substitutes. However, import substitute foods faced poor market pros-
pects in Nigeria because of lack of effective demand for nutritionally
superior goods such as eggs, milk, meat and other processed items
(Johnson, et al, 1969).
The growth rate in demand for food in Nigeria in the present study
was higher than the rate shown in the CSNRD report. There appear to be
two reasons. First, the bias, if any, may come from the nature of the
data used for income elasticities for food which were not actually
measured but projected. Second, the CSNRD study considered effective
demand as determined by the disposable income of consumers, while this
study considers effective demand as determined by population and income.
Thus, it may still be valid to conclude that Nigeria is experiencing
shortages in import substitute foods if demand is conceptualized in
terms of population and income. In fact, a simulation study conducted
in Nigeria just after the CSNRD study confirms the results of this work
(Johnson, et al, 1971, p. 298):

However, we should add that the two strategies in-
volving export crop modernization programs did
create some long-run adverse effects in the non-
agricultural sector. The increased profitability
of export crops stimulated agricultural producer

onsortium for Study of Nigerian Rural Development, a joint venture
by Michigan State University, University of Wisconsin, Ohio State Univer-
sity, Colorado State University, Kansas State University, U.S.D.A., and
the Research Triangle Institute, to work on many aspects of agricultural
and rural development in Nigeria. Nutritionally superior foods are eggs,
maize, beans, millet, grains, soybeans, meat and milk. The four major
import substitute foods are fish products, wheat, milk and cream, and
sugar. See Johnson, et al, 1969, pp. 22-14 and pp. 99-100.

Table 12.--Differences between annual rates of growth
demand for selected food items, 1965-73

of supply and

Annual rate of growth
Country Food Supply Demand Supply < Demand


Kenya Rice .177 3.40 <
Maize -.460 3.35 <
Beans .0006 3.37 <
Sugar -2.02 3.45 <

Nigeria Rice .23 2.44 <
Maize +19.96 2.42
Beans -1.29 2.44 <
Sugar 1.53 2.51 <

Tanzania Rice 1.9 2.77 <
Maize 2.937 2.75 >
Beans .24 2.77 <
Sugar 3.98 2.87 >

Zaire Rice 2.12 2.690 <
Maize .913 2.695 <
Beans .021 2.693 <
Sugar 1.42 2.683 <

Source: Computed from F.A,O. Agricultural Commodities, Projections
for 1975 and 1985, Rome 1967, and F.A.O. Production Yearbook, selected

demands for nonagricultural goods and consequently,
the nonagricultural population's demand for food.
In addition, the profitability caused some pro-
ducers to switch from food crop production to
export crop production. Consequently, the price
for food increased substantially in both programs
involving export crop modernization because food
demand increased and food supply decreased.

The above argument is illustrated in Figure 8, which also shows the
price effects of competition between the export and domestic agriculture
Growth in food demand exceeded that for supply in both Kenya and
Zaire. For Kenya, a recent report by the World Bank (Burrows) stated:
Many rural families have diets deficient in calories,
and Vitamins A, B2, and C. Although most pronounced

Super pound



.009 .. .. -
---- = Food crop modernization
--- = Only export crop modernization
....... = Export crop modernization with
.006- marketing boards take-off

.003- I ..I. I
1965 1970 Years 1975 1980

Figure 8.--Average price of food staples in northern Nigeria under alternative simulated
policies, 1965-1980.

Source: A Generalized Simulation Approach to Agricultural Sector Analysis: With
Special Reference to Nigeria, Michigan State University, East Lansing, 1971.

in Nyanza Province and parts of the eastern plateau,
and variable by season, the deficiencies are nation-
wide...Pressure on land is increasing, people are
moving into marginal areas, maize has largely taken
over from the more protein-rich millets and sorghums,
and there is no reason to believe that pulse production
is going up at a faster rate than population.
Food price increases in Zaire are evidence that increases in supply
failed to keep pace with growth in demand. Lack of data prevents any
conclusions for Tanzania.
It appears that "modernization" of export crop sectors in the study
countries, without comparable effort in food sectors, is the major
factor in the imbalance between the two. The problem, though related
to the allocation of land and labor between the two sectors, is never-
theless more far-reaching. Food and export crops compete for land to
some degree, but food crops have a much larger share of cultivated land
than do export crops (Table 13). The scientific and economic infrastruc-
ture allocated to the export sector has made it relatively more productive
and profitable than the domestic food sector. One study from Africa
(Yudelman, p. 281) and another from the Caribbean (Davis, p. 142) support
the validity of this hypothesis.

Table 13.--Land distribution among food and export crops, 1970

Country Food crops Export crops


Kenya 80 20
Nigeria 70 30
Tanzania 60 40
Zaire 78 22

Source: F.A.O., World Crop Statistics.

The two studies express strikingly similar views, as Yudelman
The philosophy of agricultural development in many
parts of Africa has been oriented toward a commodity
approach, which has required the concentration of
investments and skills on the expansion of output
of those commodities for which there is effective
demand. In most countries, these are necessarily

export commodities;...However, whatever the merits
of this approach for economic growth, it has led
to a concentration of investment in the areas or
regions with higher income producers of export crops.
This tends to widen the regional differential; farm
incomes in those regions not producing for export
tend to lag further behind as development proceeds.
Davis, in analyzing the relationship between agricultural research and
agricultural development in the Caribbean, states:
The differential level of investment between the
export and the domestic research systems raises a
further question regarding the extent to which the
differential might be related to significant dif-
ferences in the marginal value productivity of
investment. There is a reason to suspect that the
marginal value productivity of investment of the
export system was considerably higher than that of
its domestic counterpart. This suspicion is related
to the generally higher output price of export crops,
...and significantly larger marginal productivity of
research investment for the export system.
Thus, the superior performance of export crops over domestic food
crops can be explained by the capital investment allocated to the former,
in terms of modern inputs, research infrastructure and modern management.
If similar resources were also allocated to the food sector, it would
probably grow and develop at a comparable rate.8 The relatively low
growth rates of agricultural export earnings for the four countries,
though, suggest that there are less than optimum levels of investment in
the export agriculture sectors.

African Choice

This section discusses the general objective of the study, that of
identifying policies in African countries that affect the overall economy
and the export agricultural sector. It was suggested earlier that,
consciously or unconsciously, the four African countries studied have

8The authors acknowledge the importance of the point made by
P.J. van Blokland that subsistence agricultural production is persis-
tently underestimated because very little data are available. More
information is collected and available on export commodities, so that
input or "performance" of export sectors may be overstated relative
to production for on-farm or domestic consumption.

opted for higher rather than for more stable incomes. In fact, the short-
run effect of the diversification process was found to reduce export income
Evidence from this study shows that the four countries have become
more specialized in their export agricultural sectors. However, this
trend raises two questions. First, what is likely to be the impact on
the allocation of land resources between different export crops? Second,
what is the most effective use of increased agricultural export earnings
as a development tool?
The answers to these two questions were sought by using individual
country data generated by the models, and by referring to the literature
on the subject.

Effects on Export Crop Specialization on Land Allocation


Results with the two-stage least squares (TSLS) system (equations
17 and 18) show Kenya heavily dependent on the four crops described
Z 300.945 + .9732 (Qt) + .220 (Q2jt) + .153
Jt (199 ) ( 582) (.038) (.076)


+ .452 (Q t)
(.135) 4t

- 2.829 (Djt)

Zt = agricultural export earnings as defined in equation (17),
Qljt = quantity of coffee exported by Kenya, in thousands of metric
tons in period t,
Q2jt = quantity of tea exported by Kenya, in thousands of metric
tons in period t,
Q3t = quantity of cotton exported by Kenya, in thousands of metric
Q3jt tons in period t,
Q4jt = quantity of sisal exported by Kenya, in thousands of metric
tons in period t,
Djt = diversification index, as defined in equations (6) and (7),
= b-coefficient significant at the 10 percent level,
** = b-coefficient significant at the 5 percent level,
*** -= b-coefficient significant at the 1 percent level,
( ) = values in parenthesis are standard errors.

Table 14.--Export supply function estimates, Kenyaa

Crops Constant Pit-1 Pot-1 Ait Aot RFt T R2 DW

Coffee 50.435 -.250 1.562 .406 .323*** -2.741 23.43*** .920 2.338
(326) (.379) (1.287) (1.90) (.097) (5.23) (13.88)
Tea -313.73* .709 .294 1.94** 1.64 .940 -6.014 .982 1.862
(240) (1.039) (.718) (.218) (1.46) (4.413) (8.1808)
Cotton 44.837 .827 .3665 .3598 .3435 -1.204 .5554 .411 1.610
(193.76) (.780) (.654) (.464) (.800) (4.315) (.773)
Sisal 18.336 .909 -.543 1.029** -.1926 2.050 2.244 .781 1.433
(101.81) (1.32) (.306) (.451) (.291) (1.95) (3.46)

Qit = f(P Ajt

, Aot, RFt, T, u). Linear form.

= b-coefficient significant at .10
** = b-coefficient significant at .05
*** = b-coefficient significant at .01
( ) = values in parentheses are standard errors.

The four crops have different weights, as coffee has the largest
coefficient with .973, followed by sisal (.452), tea (.220) and cotton
(.153). In the 1950-73 period, more coffee was produced than other com-
modities. Export supply functions also show that the quantity of coffee
exported increased over time (Table 14). A legitimate question is whether
the increase in coffee exports affects the allocation of land to other
crops. The coefficient in the coffee export supply function (Aot) is
positive and significant, suggesting that coffee acreage moves in the
same direction as the area of other crops taken as a whole. This suggests
that the specialization in coffee production is not hindering the develop-
ment of other crops, but that those regions specializing in coffee pro-
duction are growing at a proportionally higher rate than others.


Agricultural export earnings of Nigeria are positively related to
cocoa, groundnut and rubber exports, but inversely related to palm oil
exports. Results:
** ***
Zjt o 15.90** + .416 (Qljt) + .979 (Q. ) .130 (Q3jt)
(5,24) (.130) (.115) (.055)
+ .267 (Q4jt) 2.63 (Djt
(.173) (10.64)
Zjt = agricultural export earnings as defined in equation (17),
Qljt = quantity of groundnuts exported by Nigeria, in thousands of
metric tons in period t,
Q2jt = quantity of cocoa exported by Nigeria, in thousands of metric
tons in period t,
Q3jt = quantity of palm oil exported by Nigeria, in thousands of
metric tons in period t,
Q = quantity of rubber exported by Nigeria, in thousands of
4jt metric tons in period t,
Djt = diversification index, as defined earlier,
= b-coefficients significant at the 10 percent level,
** = b-coefficients significant at the 5 percent level,
*** = b-coefficients significant at the 1 percent level,

( ) = values in parentheses are standard errors.
The magnitude of the coefficients shows that cocoa (.979) and groundnuts
(.416) have the greatest impact on the level of Nigerian agricultural

Table 15.--Export supply function estimates, Nigeriaa

Crops Constant Pi P-1 A Aot RF T R2 DW

Groundnuts -121.03 -1974** .297 .151 .776 8.752*** 3.970** .835 2.415
(138.16) (.438) (.454) (.166) (2.14) (2.46) (2.56)

Cocoa -73.747 -.4346** .6637 .9894 .2466 2.989 6.537 .785 2.905
(153) (.238) (.635) (2.61) (.301) (2.46) (3.406)
Palm'Oil -208.83** .323 -.324 1.122*** .2733** 5.361*** -8.160*** .936 1.584
(77.92) (2.09) (2.47) (.250) (.141) (1.27) (1.58)
N. Rubber -592.10*** 1.498*** -.285 .914*** .4468* 11.744 5.973 .974 2.360
(139.1) (.369) (.526) (.104) (331) (2.472) (2.666)

a Qit = f(Pit-1 Pt-i' Ait, A t, RFt, T, u).

= b-coefficient significant
= b-coefficient significant
= b-coefficient significant
= values in parentheses are

Linear form.

at .10
at .05
at .01
standard errors.

( )

export earnings. Natural rubber ranks third (.267), while palm oil
adversely affects agricultural export earnings. These results, showing
that Nigeria specialized in cocoa and groundnuts, are confirmed by the
export supply function (Table 15).
The regression coefficients of the export supply variable (Aot) for
cocoa and groundnuts are not significant, so it is not possible to say if
increases in those two exports restricted other export crops. However,
regression coefficients for rubber and palm oil are both positive and
significant. This suggests that the acreage of all four crops moved in
the same direction. The negative trend in the volume of palm oil exported
can possibly be explained by the fact that domestic consumption of palm
oil has been increasing at the same rate as population increase (Helleiner,
1965). In general the same conclusions drawn for Kenya based on different
rates of growth among export crops for various regions may be drawn for
Nigeria. The northern and southwestern regions of Nigeria have been
exporting more groundnuts and cocoa, respectively, than other regions.


For Tanzania, the results from the TSLS system shown below are not
significant, except for the coefficient related to cotton exports and
the diversification index.
Zjt = -1194.71 .513 (Qljt) + 2.53 (Q2t) + .173 (Q3jt)
(.348) (.766) (.199) (.101)
+ .670 (Q ) 9.16 (Djt)
(.680) (6.88)
Zj = agricultural export earnings as defined in equation (17),
Qljt = quantity of coffee exported by Tanzania, thousands of metric
tons in period t,

Q2jt = quantity of tea exported by Tanzania, in thousands of metric
tons in period t,
Q,3t = quantity of cotton exported by Tanzania, in thousands of
S metric tons in period t,
Q4jt = quantity of sisal exported by Tanzania, in thousands of metric
tons in period t,
Djt = diversification index as defined earlier,
b-coefficient significant at the 10 percent level,
** = b-coefficient significant at the 5 percent level,
( ) values in parentheses are standard errors.

Table 16.--Export supply function estimates, Tanzaniaa

Crops Constant Pit- Pot- Ait Aot RFt T

Coffee -223.897 -.4724 -1.2311** .2587 .6320* 9.5200* -3.7975
(268.9) (1.24) (.677) (.332) (.496) (7.40) (6.55)

Tea -596.142 -.7214 -17795** -1.0401 1.2865** 22.5307 33,1879
(805) (2.26) (104.9) (2.35) (.821) (19.5) (13.82)
Cotton 1379.28* 1.8400** ~9856 .123 3.9652** 5Q,7574** 28.8765
(950) (1.114) (1.02) (.427) (1.794) (23.12) (14.50)
Sisal 137.527* -.4194 -.4203 1.1008*** -.2423** -3.0153 6.2876**
(92.92) (.960) (1.26) (.291) (.086) (2.27) (2.01)

a Qit

( )

Linear form.

b-coefficient significant
b-coefficient significant
b-coefficient significant
values in parentheses are

at .10
at .05
at .01
standard errors.

= fP t-1 Pot-1, Ait, Aot, 11t, T, )

On the basis of these results, the only statement that may be made is
that export crops earnings depend largely on the volume of cotton produced.
The relationships between different crop areas indicate that only sisal is
affected by an increase in production of other crops as a whole (Table 16).
This may mean that sisal is competing for land with cotton, tea, and coffee


For Zaire, the logarithmic form of the TSLS equation was used.
Log Z. 1.763 + 1.465 log (Qlt)** .375 log (Q2 )
3t (5.21) (.6511) (.734) t
+ .155 log (Q3jt) + .370 log (Q ) .908 log (Dt)
(.302) (.961) 4t 1.322 jt
Z. = agricultural export earnings as defined in equation (13),
Qjt = quantity of coffee exported by Zaire, in thousands of metric
tons in period t,
Q2jt = quantity of tea exported by Zaire, in thousands of metric
tons in period t,
Q3jt = quantity of palm oil exported by Zaire, in thousands of metric
tons in period t,

q4jt = quantity of natural rubber exported by Zaire, in thousands of
S metric tons in period t,
Dt = diversification index, as defined earlier,
b-coefficient significant at the TO percent level,
= b-coefficient significant at the 5 percent level,

( ) = values in parentheses are standard errors.
Coffee is the only crop of the four with a significant coefficient.
As discussed for the other countries, the results above and in Table 17
suggest that when the coffee area increases, area in other crops as a
whole also increases. Again, the same arguments regarding the differential
between growth rates of exports of different specialized regions applies
to this case.

The Use of Agricultural Export Earnings for Development Goals

Agricultural export earnings of the four study countries increased
over the study period (Table 10). Kenya's earnings increased at 6.7 per-
cent per year, while those of Nigeria, and Tanzania grew at 4.4 and 4.1

Table 17.--Export supply function estimates, Zairea

Crops Constant Pit Pot-1 Ait Aot RFt T 2 DW

Coffee -10.804 .5429 .4178 -.630** .460** 3.692 .969*** .409 1.613
(8.69) (1.87) (.417) (.311) (.240) (2.074) (.294)

Tea 8.171 .6613* -.2237 .3116 1.135** -3.247** .9363*** .975 2.001
(7.32) (.514) (.439) (.258) (.439) (1.31) (.333)

PalmOil -3.243 -.3119** .1597 1.5901*** -.2116 .6044 -.1260 .922 1.347
(3.05) (.139) (.178) (1.62) (.207) (.598) (.138)

N. Rubber-9.548** .2011* -.1364 .7111 -.1824 3.022 .9849 .933 1.900
(4.87) (.152) (,256) (.163) (.208) (1.04) (1.93)

a Qt = ot t-1' Ait' Aot, RFt T, u).

b-coefficient significant
b-coefficient significant
b-coefficient significant
values in parentheses are

Double logarithmic form.

at .10
at .05
at .01
standard errors.

( )

percent respectively. Zaire experienced a much slower rate of growth in
export earnings with 0.3 percent annually (Table 10). If export earnings
.grow at low rates and agricultural exports constitute a relatively large
share of GDP, there is reason to expect that the rate of overall develop-
ment will be slow.
Several authors have attempted to explain why the dynamics of the
export economy have not broken the vicious circle of poverty in LDCs
(Singer, Prebisch and Levin). Singer and Prebisch blamed the specialized
nature and structural patterns of export economies. Levin argued that
the structure of export economies and their isolation from national
markets is the basis of the development handicap. Johnson and others
(1971) suggested that agricultural export earnings surpluses gained by
marketing boards were not used in the agricultural sector to meet develop-
ment goals. For Nigeria, however, this view is strongly opposed (Aboyade,
Helleiner, 1966), and in Tanzania, Kriesel believed agricultural export
earnings were used effectively in the agricultural sector.
The findings of this study suggest that the performance and returns
of marketing board investments in development projects have been poor.
There are two reasons for this conclusion: First, the apparent lack of
responsiveness of marketing boards to changes in world prices (Tables
14-17); and second, relatively low annual rates of growth in export ear-
nings (Table 10 and Figure 2), despite the use of substantial land
resources for export crop production.
Lack of responsiveness of marketing boards to world prices may be
explained by three factors. The first is pricing policies which kept
producer prices well below world prices, removing incentives to increased
production among peasant farmers. The second factor relates to the lack
of technological investment in agriculture, and the fixity of factors of
production. Fixity of factors means that farmers do not have a wide
choice of input combinations, but instead, rely heavily on land and labor.
Factors such as fertilizers and capital are used little, partially because
farmers have incomes too low to afford them. Marketing board surpluses
passed on to farmers would enable them to acquire inputs that would in-
crease production. However, such surpluses have been siphoned off for
"political parties" (Nixon) and for "dubious projects" (Eicher, 1970)
in many cases. The third factor is the perennial nature of many export
crops and the resulting restrictions on the degree of flexibility in
earning use.

There is evidence that although agricultural export earnings have
historically low growth rates, the agricultural export sector is still
an appropriate development tool. These low rates appear to be caused
primarily by organizational and structural problems. If these problems
were removed, the export crop sector could stimulate overall agricultural
development. These structural problems, although pronounced in countries
with marketing boards, are also present in other nations. In the latter
countries, the problems are largely related to direct government fiscal
and pricing policies affecting the farm sector. Thus, rather than accuse
the export sector of hindering the development process, efforts should be
made to stimulate expansion of the overall agricultural sector, without
reducing the export sector (McPherson, 1974). But, whether such an
expansion should be based on agricultural diversification policies is
still not answered with the results reported here.
The findings suggest, however, that there are some positive effects
accruing to diversification. Furthermore, the findings indicate that
specialization in certain commodities is due to the fact that certain
regions are expanding their exportable supplies of those commodities
proportionally more than others. Thus, diversification can be a viable
policy if used to stimulate export crop production in economically re-
tarded areas. On the other hand, diversification can have a retarding
influence if it involves the use of export crops in developing areas to
achieve sub-optimal levels of agricultural diversification in the hope
of achieving economic development.



This study focused on four African countries: Kenya, Nigeria,
Tanzania and Zaire. The objectives were (a) to identify whether agri-
cultural export earnings fluctuations were determined primarily by
quantity or price variability, (b) to investigate trends and patterns in
the agricultural export sector, (c) to inquire into the relationships
between agricultural export earnings fluctuations and the export crop
diversification process and (d) to determine the difference between food
supply and food demand growth rates.

Fluctuations in agricultural export earnings were found to be due
mainly to variations in quantities exported, rather than to price vari-
ability. These quantity variations appear to be related to such factors
as the lack of flexibility by institutions responsible for the trading
of such export crops, and agro-ecological variability. These institutions
have continuously exported the traditional agricultural commodities.
Such a policy demonstrates a lack of responsiveness to world prices. A
secondary source of export quantity variation was variability in rainfall
and its effects on crop production.
Results of the empirical analysis showed that all four countries,
although diversified in terns of the number of export crops produced in
different ecological regions, have been specializing in one or two export
crops. The pattern of such specialization is not characterized by com-
petition between crops as such, but rather, by unequal growth rates of
specialized production regions.
Results from analyzing the relationships between export crop diversi-
fication and agricultural export earnings revealed an inverse relationship
between these two variables. That is, an attempt to increase diversifi-
cation would result in reducing the level of agricultural export earnings,
at least in the short run. However, results from the quadratic model
reveal that an intensified diversification program may cause the level of
agricultural export earnings to rise in the long run.
While the results generated by the models aimed at analyzing the
effects of diversification on export earnings fluctuations were incon-
clusive, they did indicate that the decrease in diversification since
1950 was accompanied by a higher degree of instability in some countries
and by an increase in earnings stability in other countries.
In regard to the difference between growth rates of food supply and
food demand, it was found that demand for food was increasing more ra-
pidly than food supply in all four countries. The argument based on
the land allocation to these two sectors, in conjunction with findings
of other authors, suggest that the gap between the two sectors is due to
a relatively mild advantage of export crop over food crops, in terms of
investment in capital and research infrastructure. Such investments,
although sub-optimal in the export crop sector itself, nevertheless have
made this sector relatively more "modern" than the domestic food pro-
duction sector.

For the general objective of the study, an attempt was made to de-
termine if specialization inhibited the development of other crops by
reallocating land resources. In general, the areas of leading crops
were moving in the same direction as those of lagging crops. One possible
explanation is that crops were grown separately in specific ecological
regions and some regions were more committed than others to the growing
of these crops. Another aspect explored in the study relates to the
low rates of growth of agricultural export earnings under a specialization
policy. It was argued that these slow growth rates are due to the lack
of an appropriate level of modernization in the export crop sector.
Specialization based mainly on the increased use of only land and labor
was held responsible for such low performance.


Since it appears that a large proportion of variability in export
earnings is due to fluctuations in quantities, it is then imperative to
use policies that address themselves more to quantity adjustment than to
price regulation. Such policies could be buffer stocks and quota adjust-
ments. However, because of difficulties inherent in international policy
enforcement (Edwards and Parikh), emphasis should be given to alternative
domestic adjustments such as export crop diversification and appropriate
tax structures. These policies would be most appropriate for those
countries which have specialized in crops whose variability in quantity
is higher. Also, because of the long-run effects of export crop diversi-
fication on the level of export earnings, diversification should be con-
sidered. But since the effects on instability of this policy would vary
according to the country and the physical and economic characteristics of
crops introduced, such a policy would have to be adopted cautiously.
The need for caution is apparent when one considers the cases of
Tanzania and Zaire. In Tanzania, further diversification would result
in higher instability, while in Zaire it would stabilize earnings (equation
15). The results from these countries lead to the conclusion that there
exists a range where export diversification has desirable effects but
works against stability and higher incomes beyond that range. These
ranges would be expected to differ from country to country. Further
research, to determine conditions and causes of patterns in different

countries would be interesting and useful.
The gap between the domestic food sector and the export crop sector
reinforces the feeling that the lack of modernization of the food sector
is responsible for such an imbalance. However, the balanced modernization
of both export and food sectors would help the former to grow at a faster
pace and at the same time, help the latter sector to keep up with demand
Another conclusion regarding the general objective of the study is
related to the finding that there exist unequal growth patterns of export
crop production among regions within a country. Such a situation may be
the source of the observed specialization trend and the cause of a failure
to achieve optimal diversification levels. It is believed that efforts
to stimulate the export production of economically retarded regions would
not only help a country increase its export crop supplies but also help
to achieve an optimum degree of diversification. Such an effort requires
a combined policy of additional investments in the agricultural of re-
tarded areas and relevant pricing system at the producer level.


Table 18.--Average annual values in current dollars of world and
country exports of major commodities, 1950-73

World Kenya Nigeria Tanzania Zaire
Commodity Value Value % Value % Value % Value %
$1,000,000 $1,000,000 $1,000,000 $1,000,000 $1,000,000

Coffee 2,552.8 35.6 1.4 29.6 1.2 41.3 1.6
Tea 630.9 20.6 3.3 -- 4.2 0.7 1.9 0.3
Cocoa 585.9 -- -- 112.1 19.1 -- -- -- -
Cotton 2,307.5 2.0 0.1 -- 26.4 1.1 -- -
Sisal 119.4 10.5 8.8 -- -- 40.5 33.9 --
Rubber 1,494.2 -- -- 25.0 1.7 -- -- 15.6 1.0
Groundnuts 5,063.9 -- 79.8 1.6 -- -
Palm Oil 163.0 -- 33.5 20.5 -- 32.9 20.2

Table 19.--Earnings from major agricultural exports, and total agricul

tural exports, current dollars,


Year Coffee Tea Cotton Sisal Total Total Ag. Exp. Earnings
-------------Thousand U.S. Dollars--------------------- ------Million U.S. Dollars----

















Year Cocoa Groundnuts Palm Oil Rubber Total Total Ag. Exp. Earnings

-------------Thousand U.S. Dollars----------------------- ------Million U.S. Dollars-

1950 64,292 76,515 63,378 8,760 212,945
1951 95,463 32,915 87,522 27,614 243,514 308.7
1952 92,001 63,566 62,933 15,824 234,324 287.6
1953 81,766 81,222 68,301 11,521 282,810 295.0
1955 71,647 103,993 65,262 26,629 267,536
1956 66,880 74,665 29,246 170,971 324.5
1957 79,475 64,205 37,827 24,116 205,623 300.6
1958 74,938 91,245 34,995 21,878 223,056 331.7
1959 107,210 76,920 38,663 32,490 255,283 417.9
1960 98,159 61,475 36,907 39,880 236,421 403.0
1961 94,490 90,251 37,035 30,730 252,506 416.5
1962 93,372 90,793 24,998 31,490 240,653 356.0
1963 90,605 102,463 26,222 32,750 252,040 359.7
1964 112,279 95,920 30,111 33,850 272,160 419.5
1965 119,534 105,854 38,055 30,590 294,033 463.4
1966 79,129 114,280 30,697 32,020 256,126 433.0
1967 153,127 99,156 3,527 17,770 273,580 412.1
1968 144,874 106,628 39,900 17,670 309,072 405.3
1969 147,269 100,271 121,300 26,919 395,759 426.8
1970 186,305 60,842 159,000 24,596 430,743 438.6
1971 201,790 34,019 477,800 17,492 731,101 392.4
1972 153,724 32,528 41,800 12,495 240,547 315.2
1973 170,796 69,169 1,400 29,482 270,847 456.5


Table 19.--Earnings from major agricultural exports, and total agricultural exports, current dollars,

Year Coffee Tea Cotton Sisal Total Total Ag. Exp. Earnings

-------------Thousand U. S. Dollars------------------- -----Million U.S. Dollars----

1950 14,956 574 7,903 50,336 73,769
1951 21,735 7,387 94,523 123,645
1952 23,743 864 8,792 72,073 105,472
1953 20,429 1,193 11,373 47,226 80,221
1955 26,767 2,353 16,196 40,457 85,773
1956 35,881 2,336 20,895 42,399 101,511 109.5
1957 19,420 2,625 17,747 26,966 66,758 92.8.
1958 20,702 2,839 20,962 28,786 73,289 97.2
1959 16,080 2,157 18,640 36,559 73,436 105.9.
1960 20,510 3,221 24,710 43,236 91,677 128.6
1961 18,930 3,742 19,020 39,278 80,970 112.1
1962 18,410 4,513 20,700 44,056 87,679 120.6
1963 19,150 4,346 30,010 63,479 116,985 156.1
1964 30,940 4,367 27,670 61,227 124,204 168.5
1965 24,060 4,230 34,190 39,989 102,469 145.8
1966 43,400 6,321 48,990 32,855 130,566 204.4
1967 33,440 6,194 35,190 28,191 103,015 174.2
1968 37,152 6,612 39,653 22,304 105,721 173.8
1969 36,104 6,863 32,886 22,352 98,205 185.3
1970 43,746 5,988 34,616 25,038 109,391 193.4
1971 32,059 5,812 34,409 18,985 91,265 190.9
1972 53,663 7,540 47,172 19,424 127,799 235.6
1973 70,591 7,734 47,767 31,786 157,873 270.2


...... .. ,.. .ujt 4UUJUI u L y1 u.uLtural i ApUI'Ls, dlna tLaI adricuiLurat exporIL, curreInL 001U 15.,

Year Coffee Tea Palm Oil Rubber Total Total Ag. Exp. Earnings

-----------Thousand U. S. Dollars------------------------ -------Million U.S. Dollars--

1950 36,240 45,667 6,113 88,020
1951 44,634 73,692 15,891 134,217 191.3
1952 38,819 86 50,911 14,293 104,109 151.4
1953 44,680 217 44,155 9,654 98,706 134.2
1955 62,219 1,115 52,543 22,492 138,369
1956 83,342 1,519 59,959 24,497 169,317
1957 68,178 2,267 34,503 20,374 125,322
1958 63,570 2,958 32,922 20,227 119,677
1959 61,560 2,517 38,261 22,320 124,658
1960 30,210 2,957 30,322 25,840 89,329
1961 23,700 16 31,894 21,490 77,100
1962 14,520 2,268 28,246 20,170 65,204
1963 17,830 2,416 23,570 17,070 60,886
1964 24,190 1,440 21,190 13,630 60,450
1965 14,740 2,011 18,500 9,150 44,401
1966 23,050 2,213 15,450 10,480 51,193
1967 28,060 2,900 21,700 9,610 62,270
1968 34,300 3,500 24,800 9,900 72,500
1969 28,200 1,305 19,269 16,273 65,047
1970 33,900 1,892 28,181 12,759 76,732
1971 41,000 1,300 26,404 12,000 80,704
1972 67,251 2,544 16,600 10,500 96,895
1973 65,000 1,470 17,500 14,000 97,970

Source: F.A.O.,

Trade Yearbook .

Table 20.--Values of commodities exported, current dollars, 1950-1973

Coffee Tea Cocoa
Year Kenya Tanzania Zaire World Kenya Tanzania Zaire World Nigeria World
----------Thousand------- -Million- -----Thousand----------- i Million- -Thousand- -Million-

1950 11,505 14,956 36,240 2,220.4 3,560 574 471.9 64,292 450.1
1951 12,937 21,735 44,634 2,496.9 95,463 553.1
1952 21,357 23,743 38,819 2,449.7 3,715 864 86 380.1 92,001 516.8
1953 19,770 20,429 44,680 2,754.6 3,253 1,193 217 534.6 81,766 530.7
1955 28,0-9 26,767 52,219 2,954.4 6,069 2,?53 1,115 548.7 71,647 575.2
1956 44,199 35,881 83,342 3,532.5 8,293 2,336 1,519 605.1 .56,2 30 430.5
1957 23,346 19,420 68,178 2,309.8 8,709 2,625 2,267 597.7 79,475 464.1
1958 23,266 20,702 63,570 2,004.2 9,819 2,239 2,958 635.3 74,933 554.9
1959 29,660 16,080 61,550 1,970.9 10,212 2,157 2,517 604.9 107,210 570.1
1960 23,770 20,510 30,210 1,912.1 12,465 3,221 2,957 611.7 98,159 542.6
1961 29,740 18,930 23,700 1,859.2 11,851 3,742 16 626.5 94,490 487.4
1962 29,720 18,410 14,520 1,886.2 15,609 4,513 2,268 664.1 93,372 473.2
1963 30,890 19,150 17,830 1,995.0 20.449 4,346 2,416 679.9 90,605 507.9
1964 43,160 30,940 24,190 2,387.4 21,446 4,367 1,440 675.9 112,279 522.5
1965 39,510 24,060 14,740 2,231.3 21,669 4,230 2,011 684.9 119,534 501.6
1966 52,610 42,400 23,050 2,393.9 29,131 6,321 2,213 632.4 79,129 457.0
1967 43,920 33,440 28,060 2,239.9 25,520 6,194 2,900 730.7 153,127 593.0
1968 35,976 37,152 34,300 2,555.4 32,137 6,612 3,500 708.2 144,874 641.7
1969 47,268 36,104 28,200 2,490.4 34,683 6,863 1,305 625.1 147,269 795.6
1970 62,428 43,746 33,900 3,081.9 40,183 5,988 1,892 696.0 186,305 867.6
1971 54,777 32,059 41,000 2,748.6 34,085 5,812 1,300 693.8 201,790 737.0
1972 69,406 53,663 67,251 3,227.7 49,668 7,540 2,544 724.1 153,724 707.2
1973 70,591 65,000 4,321.5 51,129 7,734 1,470 747.9 170,796 944.8

Average 35,557.4 29,602.96 41,269.26 2,552.78 20,620.6 4,201.09 1,352.90 630.89 112,135.87 585.85

VI U. u

Cotton Groundnuts Sisal
Year Kenya Tanzania World Nigeria World Kenya Tanzania World
---Thousand---- ---Million-- -Thousand- -Million- ----Thousand------- --Million--

1950 952 7,903 2,256.5 76,515 199.4 13,633 50,336 147.2
1951 1,583 7,387 2,163.3 32,915 123.8 25,791 94,523 284.7
1952 2,101 8,792 1,812.8 63,566 153.8 16,096 72,073 173.1
1953 1,441 11,373 1,849.9 81,222 168.8 8,811 47,226 112.1
1954 103,998
1955 2,113 16,196 1,842.6 232.3 7,793 40,457 115.5
1956 2,289 20,895 2,093.2 64,205 8,048 42,399 116.5
1957 900 17,747 1,967.6 91,245 218.8 5,869 26,966 78.3
1958 1,415 20,962 1,716.2 76,920 227.2 6,120 28,786 81.7
1959 1,840 18,640 1,844.4 61,475 202.2 9,683 36,559 102.3
1960 2,350 24,710 2,441.1 90,251 193.2 12,784 43,236 118.2
1961 1,760 19,020 2,351.3 90,793 242.0 11,737 39,278 112.4
1962 1,220 20,700 2,053.8 102,463 245.8 12,105 44,056 132.6
1963 1,220 30,010 2,257.0 95,920 256.0 21,090 63,479 185.3
1964 1,810 27,670 2,372.1 105,854 264.6 16,847 61,227 173.3
1965 2,090 34,190 2,295.4 114,280 272.1 10,785 39,989 113.4
1966 2,430 48,990 2,307.0 99,156 291.0 9,352 32,855 101.1
1967 1,760 35,190 2,237.6 106,628 246.8 5,833 28,191 77.6
1968 1,115 39,653 2,375.4 100,271 258.9 5,157 22,304 72.1
1969 2,131 32,886 2,296.2 60,842 254.9 4,821 22,352 73.5
1970 3,437 34,616 2,484.1 34,019 216.0 5,258 25,038 73.7
1971 3,310 34,409 2,796.3 32,528 208.2 4,264 18,985 65.0
1972 3,410 47,172 3,133.1 69,169 236.0 5,801 19,424 79.4
1973 3,913 47,767 4,125.9 334.1 13,671 31,786 158.0

Average 2,025.6 26,386.0 2,307.51 79,737.95 5,063.90 10,536.91 40,501.09 119.43


Table 20.--Values of commodities exported, current dollars, 1950-1973--Continued

Palm Oil Rubber
Year Nigeria Zaire World Nigeria Zaire World
-----Thousand----- --Million-- -----Thousand------ ----Million--

1950 63,378 45,667 182.3 8,760 6,113 1,815.8
1951 87,522 73,692 281.9 27,614 15,891 3,282.5
1952 62,933 50,911 188.8 15,824 14,293 1,897.2
1953 68,301 44,155 191.0 11,521 9,654 1,066.7
1955 65,262 52,543 192.6 26,629 22,492 1,920.9
1956 74,665 59,959 149.4 29,246 24,497 1,773.9
1957 37,827 34,503 79.4 24,116 20,374 1,387.0
1958 34,995 32,922 74.4 21,878 20,227 1,277.6
1959 38,663 38,261 122.5 32,490 22,320 1,725.7
1960 36,907 30,322 118.7 39,880 25,840 1,804.6
1961 37,035 31,894 122.4 30,730 21,490 1,399.2
1962 24,998 28,246 105.3 31,490 20,170 1,528.8
1963 26,222 23,570 113.1 32,750 17,070 1,400.4
1964 30,111 21,190 127.2 33,850 13,630 1,244.8
1965 38,055 18,500 146.8 30,590 9,150 1,281.5
1966 30,697 15,540 143.3 32,020 10,480 1,306.0
1967 3,527 21,700 112.0 17,770 9,610 1,222.7
1968 399 24,800 109.9 17,670 9,900 890.3
1969 1,213 19,269 123.5 26,919 16,273 1,246.7
1970 1,590 28,181 200.9 24,596 12,759 1,126.7
1971 4,778 26,404 281.5 17,492 12,000 965.5
1972 418 16,600 261.7 12,495 10,500 891.4
1973 14 17,500 382.4 29,482 14,000 1,910.1

Average 33,456.96 32,879.96 162.96 15,597.09 1,494.17 25,039.22


Abbott, J. C. "Agricultural Marketing Boards in Developing Countries,"
Journal of Farm Economics, Vol. 49, 1967, pp. 705-722.

Aboyade, 0. "A Note on External Trade, Capital Distortion and Planned
Development," in I. G. Stewart, African Primary Products and
International Trade, University Press, Edinburgh, 1965, pp. 26-43.

Agency for International Development. A.I.D. Economics Data Book:
Africa, Statistics and Reports Division, March 1971.
Aigner, D. J. Basic Econometrics, Prentice-Hall, New Jersey, 1971.

Balassa, B. Trade Prospects for Developing Countries, The Economic
Growth Center, Yale University, 1964.

Beckford, G. L. Persistent Poverty: Underdevelopment in Plantation
Economies of the Third World, Oxford University Press, New York,

Bieber, J. Diversification Opportunities and Effects of Alternative
Policies on Costa Rican Coffee Farms, Ph.D. Dissertation, University
of Florida, 1970.
Burrows, J. Kenya: Into the Second Decade, Report of a mission sent to
Kenya by the World Bank, The Johns Hopkins University Press,
Baltimore and London, 1976.

Carter, H., G. W. Dean and A. D. Reed. Risk and Diversification for
California Crops, California Experiment Station, Berkeley,
Circular #503, 961.

Committee for Economic Development. Trade Policy Toward Low Income
Countries, Report, 1967.

Coppock, J. D. International Economic Instability, McGraw-Hill Book
Company, New York, 1962.

Foreign Trade of the Middle East, Economic Research
Institute, American University of Beirut, Beirut, 1966.

Davis, C. G. "Agricultural Research and Agricultural Development in Small
Plantation Economies: The Case of the West Indies," Social and
Economic Studies, Vol. 24, No. 1, March 1975, pp. 117-152.

De Vries, B. A. The Export Experience of Developing Countries, IBRD,
The John Hopkins University Press, Baltimore, 1967.

De Wilde, J. C. Experiences with Agricultural Development in Tropical
Africa, Vol. 1: The Synthesis, Vol. 2: The Case Studies, The Johns:
Hopkins University Press, Baltimore, 1967.

Draper, N. R. and H.
Sons, Inc., New

Smith. Applied Regression Analysis, John Wiley &
York, 1966.

Edwards, R. and A. Parikh. "A Stochastic Model Policy Simulation of the
World Coffee Economy," American Journal of Agricultural Economics,
May 1976, pp. 152-160.

Eicher, C. and C. Liedholm.
Economy, Michigan State

Growth and Development of the Nigerian
University Press, East Lansing, 1971.

Research on Agricultural Development in
Five English-Speaking Countries in West Africa, The Agricultural
Development Council, New York, 1970.

Finke, J. and E. R. Swanson. "Diversification in Illinois Crop Produc-
tion, 1938-70," Illinois Agricultural Economics, January 1973.

Fitzgerald, D. A., S. R. Stewart, S. L. Willet, N
Bethke. Food Grain Production and Marketing
Checci and Company, Washington, D. C., 1970.

. Pritchard and S. F.
in West Africa, A.I.D.,

Flores, X. Agricultural Organizations and Development, International
Labor Office, Geneva, 1970.

Food and Agricultural Organization.
for 1975 and 1985, Vol. 2, Rome,

Agricultural Commodities--Projectionjs

* Production Yearbook.

. World Crop Statistics, Rome, 1966.

. Trade Yearbook.

Gandolfo, G.

Mathematical Methods and Models in Economic Dynamics, North
Co., Amsterdam, 1971.

Goldberger, A. S. Topics in Regression Analysis, MacMillan Company, 1968.

Heady, E. 0. "Diversification in Resource Allocation and Minimization
of Income Variability," Journal of Farm Economics, Vol. 34, 1952.

Heady, E. 0. and L. R. Whiting. Externalities in Transformation
of Agriculture: Distribution of Benefits and Costs from Development,

Iowa State University Press, Ames, 1975.

Helleiner, G. K. "Peasant Agriculture Development and Export Instability:
The Nigerian Case," in I. G. Stewart, African Primary Products and
International Trade, University Press, Edinburgh, 1965.

Peasant Agriculture, Government and Economic Growth
in Nigeria, Homewood, Illinois, Richard Irwin, 1966.

I- '

International Monetary Fund. Survey of African Economies, Vol. 2, 4,
Washington, D. C., 1969, 1971.

Johnson, G. L. "Factor Markets and Development," in W. W. McPherson,
Economic Development of Tropical Agriculture, University of Florida
Press, Gainesville, 1968.

Johnson, G. L., 0. J. Schovile, G. K. Dike and C. K. Eicher. Strategies
and Recommendations for Nigerian Rural Development, 1969-85, East
Lansing: Consortium for the Study of Nigerian Rural Development,
Michigan State University, 1969.

Johnson, G. L., M. L. Hayenga, A. N. Halter, T. W. Carrol, M. H. Abkin,
D. R. Byerlee, K. Chong, G. Page and E. Kellog. A Generalized
Simulation Approach to Agricultural Sector Analysis: With Special
Reference to Nigeria, Michigan State University, East Lansing,
November 1971.

Johnston, B. F. The Staple Food Economies of Western Tropical Africa,
Stanford University Press, Stanford, 1958.

Johnston, B. F. and J. W. Mellor. "The Role of Agriculture in Economic
Development," American Economic Review, September 1961, pp. 571-581.

Jorberg, L. Growth and Fluctuations of Swedish Industry, 1869-1912,
Almqvist & Wicksen, Stockholm, 1961.

Kenya. Statistical Abstract, Central Bureau of Statistics, Ministry of
Finance and Planning, 1960-1970.

Kingston, J. L. Instability of Export Proceeds of Selected Latin American
Countries, Ph.D. Dissertation, Pennsylvania State University, 1969.

Kriesel, H. C., C. K. Laurent, C. Halpern, and H. E. Larzelere. Agricul-
tural Marketing in Tanzania, Background Research and Policy Proposals,
Michigan State University, Agricultural Economics Dept., June 1970.

Levin, J. V. The Export Economies, Their Patterns of Development in
Historical Perspective, Harvard University Press, Cambridge, 1960.

Lim, D. "Export Instability and Economic Development: The Example of
West Malaysia," Oxford Economic Papers, Vol. 26, March 1974, No. 1,
pp. 78-92.

Mabala, K. M. L'impact de 1'Emploi et du Revenue sur l'Exode Rural dans
la region de Bandundu, Memoire de Licence en Sciences Economiques,
University Nationale du Zaire, Juillet, 1973.

Maizels, A. Industrial Growth and World Trade, Cambridge University Press,
New York, 1963.

"Export Instability and Economic Development: A Review
Article," American Economic Review, Vol. 58, No. 3, 1968.

Massel, B. F. "Export Concentration and Fluctuations in Export Earnings:
A Cross-Sectional Analysis," American Economic Review, Vol. 54, 1964,
pp. 47-63.

"Price Stabilization and Welfare," Quarterly Journal of
Economics, Vol. 83, No. 2, May 1969, pp. 284-298.

"Some Welfare Implications of International Price Stabil-
ization," Journal of Political Economy, Vol. 78, No. 2, March-April 1970.

McPherson, W. W. and B. F. Johnston. "Distinctive Features of Agricul-
tural Development in the Tropics," in B. F. Johnston and H. M. South
worth, Agricultural Development and Economic Growth, Cornell Univer-
sity Press, Ithaca, New York, 1967.

McPherson, W. W. Economic Development of Tropical Agriculture, Universit
of Florida Press, Gainesville, 1968.
"Role of Agricultural Trade in Economic Development,"
Rivista di Agricoltura Subtropicale e Tropicale, Vol. 68, 1974,
pp. 59-78.

Michaely, M. Concentration in International Trade, North-Holland Company
Amsterdam, 1962.

Mintz, I. Cyclical Fluctuations in Exports of the United States Since 18 9
NBER, New York, 1967.

Motha, G., T. C. Sheales and M. M. Saad. "Fluctuations in Australian
Rural Production and Prices, Some Implications for Support Policies,'
Quarterly Review of Agricultural Economics, Vol. 28, No. 1, January

Myint, H. The Economics of Developing Countries, Praeger, New York, 1965.

Nigeria. Annual Abstract of Statistics, Federal Office of Statistics,

Nixon, C. R. "The Role of the Marketing Boards in Political Evolution
of the Nigerian Economy," in Eicher and Liedholm, Growth and Develop-
ment of the Nigerian Economy, Michigan State University Press,
East Lansing, 1971.

Norman, D. W. "Labour Inputs of Farmers: A Case Study of the Zaria
Province of the North-Central State of Nigeria," The Nigerian
Journal of Economic and Social Studies, Vol. 11, No. 13-14, 1969.

Nurkse, R. "Trade Fluctuations and Buffer Policies of Low Income Coun-
tries," Kyklos, Vol. 11, 1958, pp. 141-230.

Ohkawa, K. "Economic Growth and Agriculture," Annals of Hitotsubashi
Academy, October 1956, pp. 46-60.

Parikh, A. "A Model of the World Coffee Economy, 1950-1968," Applied
Economics, Vol. 6, 1974, pp. 23-43.

Peemans, J. P. "The Social and Economic Development of Zaire Since Inde-
pendence: An Historical Outline," African Affairs, Vol. 74, No. 295,
April 1975, pp. 148-149.

Prebisch, R. "Commercial Policy in the Underdeveloped Countries,"
American Economic Review, Vol. 49, May 1959, pp. 251-273.

Rowe, J. W. F. Primary Commodities in International Trade, Cambridge
University Press, Cambridge, 1965.

Saint-Marc, M. Commerce Exterieur de Development, Le Cas de la Zone
Franc, Societe d'Edition d'Enswignement Superieur, 1968.

Schultz, T. W. Transforming Traditional Agriculture, Yale University
Press, New Haven and London, 1964.

Singer, H. W. "The Distribution of Gains Between Investing and Borrowing
Countries," American Economic Review, Vol. 40, May 1950, pp. 473-485,

Stewart, I. G. African Primary Products and International Trade,
University Press, Edinburgh, 1965.

Tanzania. Taarifa ya Tarakimu, Vol. 21, No. 3, 1971.

Tims, W. Nigeria: Options for Long-Term Development, The Johns Hopkins
University Press, Baltimore and London, 1974.

United Nations. Instability in Export Markets of Underdeveloped Countries,
New York, 1952.

Survey of Economic Conditions in Africa, New York, 1971.
Yudelman, M. "Distribution of the Benefits and Costs of Development in
Africa," Externalities in the Transformation of Agriculture: Distri-
bution of Benefits and Costs from Development, Iowa State University
Press, Ames, 1975.

Zaire (Republique Democratique du Congo). Rapport Annuel de la Banque
National, 1969.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs