“HE“

RSITY LIBRARIE

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII II1IIIIIIIIl

312930089

 

 

This is to certify that the
thesis entitled

NON-FARM EMPLOYMENT AMDNG FARM HOUSEHOIDS IN MNDA

presented by

Kanpayana Theobald

has been accepted towards fulfillment
of the requirements for

NH degree in $020101. (JG-y

W7

Major professor

Date g/VOV /‘I‘)c) DWIEL C. (3:,4/V

0.7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
MlchIgan State
UnIversIty

 

 

 

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

DATE DUE DATE DUE DATE DUE

E L
I;____=____

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fir:

usu Is An AflirmetIve Action/Equal Opportunity Institution
cMMflfi-D.‘

 

 

 

 

NON-FARM EMPLOYMENT AMONG FARM HOUSEHOLDS IN RWANDA

BY

Kampayana Theobald

A THESIS

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

MASTER OF ARTS

Department of Sociology

1990

(A
V

wa— ‘%

ABSTRACT

NON-FARM EMPLOYMENT AMONG FARM HOUSEHOLDS IN RWANDA
BY
Kampayana Theobald

This study examines the structure of non-farm
employment in Rwanda. Special attention is given to the
situational circumstances under which households look to
such employment as a matter of necessity and/or as an
alternative to traditional work roles. Additionally, this
study explores the importance of non-farm employment
opportunities vis-a-vis the inequalities in land and income
distribution. Based on data from a recent survey of 1,019
farm households, major differences are observed in non-farm
participation by farm size group, education level and other
household and individual characteristics. Households that
have managed to secure both land resources and the skills
necessary to participate in the non-farm sector put
themselves in an exceptional position vis-a-vis the
generation of income and accumulation of wealth. At the
other extreme, are the near-landless farm households with
little formal education and no training for alternative
employment. These households, underemployed on their own
farms have begun to rely on the meager wages they earn as

day laborers on the farms of their neighbors.

ACKNOWLEDGEMENTS

The conceptual underpinnings and empirical basis for
this research have emerged from a great deal of
collaborative effort. Without the constant guidance of my
advisor, Dr. Dan Clay, who coordinated work on non-farm
strategies in Rwanda, I would never have been able to
explore this interesting topic. All my thanks go to Dan for
his needed help and advice. I would like also to ackowledge
the members of my thesis committee, Dr. Craig Harris and Dr.
Harry Schwalzweller, for their suggestions and helpful
advice without which important pieces of this thesis
undoubtedly would have been overlooked.

Major contributions to the study design and
implementations have also been made by members of the DSA
(Division des Statistiques Agricoles, Ministry of
Agriculture) team. In particular I wish to recognize Mr.
Kayitsinga Jean and short term consultant Catherine Tardif-
Douglin who finalized the study questionnaire while I was
away for training at MSU. Our permanent team of supervisors
in each prefecture of the country, enumerators who collected
data, and those who checked, double-checked and computerized

the survey also deserve a strong word of recognition.

iii

My deepest thanks also go to the staff of the Division
of Statistics of MINAGRI, and to USAID who provided the
necessary funding for this research and for my program of
study at MSU. Last but not least, I would like to thank my
cherished parents, with whom I will not have an opportunity
to share this memorable event, my brothers and sisters and
all my friends at home, in the U.S., and elsewhere, for
their strong moral support expressed throughout in the form
of personal messages and gifts. My warmest appreciation

extends to all of you for all that you have given me.

iv

TABLE OF CONTENTS

LIST OF TABLES O O O O O O O O O O O O O O O O O O O 0 Vi

LIST OF FIGURES O O O O O O O O O O O O O O O O O O O 0 Vi i

I NTRODUCT I ON C O O O O O O I O O O O O O O O O O O O O 1
NON-FARM EMPLOYMENT: CONCEPTS AND LITERATURE . . . . . 3
Non-farm Employment in the Rwandan Context . . . . 11
Structure of employment . . . . . . . . . . . . . 13
Non-farm Employment and Income Distribution . . . 16

HISTORICAL AND SOCIO-CULTURAL CONTEXT . . . . . . . . . 17
DATA AND METHOD 0 O O O C O O I O O O O O O O O O O O O 2 2

FINDINGS AND DISCUSSION . . . . . . . . . . . . . . . 23
Patterns of Off- farm Employment . . . . . . . . 23
Non-farm Employment and Characteristics of

Individuals . . . . . . . . . . . . . . . . 27
Multivariate analysis of Agricultural Wage

Labor and Non-farm Employment . . . . . . . . 30
Focus on High Income Households . . . . . . . . . 33
Non-farm Labor and Inequality . . . . . . . . . . 35

CONCLUSIONS 0 I O O O O O O O O O O O O O O O O O O O O 3 8

LIST OF REFERENCES . . . . . . . . . . . . . . . . . . 42

Table

Table

Table

Table

Table

Table

LIST OF TABLES

Percentage of Days Worked in the non-farm
sector by type of Employment . . . . . .

Days of non-farm work by type of employment

and farm size . . . . . . . . . . . . . .
Percent of of non-farm employment and mean
days worked per year by characteristics of
indiViduals O O O O O O O O O O O O O O 0

Multiple regression of days of non-farm

activities by household members in the non-

agricultural sector and in agricultural
sector, and days of wage labor hired by
househOIds 0 O O O O O O O O O O O O I 0

Percent of households, total income, on-farm

and non-farm by income group . . . . . .

Percentage of households, total income,
on-farm and non-farm by farm size group .

vi

24

26

28

31

36

37

LIST OF FIGURES

Figure 1. Map of Rwanda . . . . . . .

vii

INTRODUCTION

Despite the importance of non-farm employment as a
means for supplementing the meager incomes of farm
households throughout the Third World, the subject has
received little systematic attention from social scientists
in Rwanda. This situation is particularly disturbing given
the country’s limited natural resources and rapidly growing
population, a problem observed as early as 1949 when
population density in Rwanda was estimated to be one of the
highest in Africa, and much higher than was found in the
neighboring countries of Tanzania, Uganda and Zaire (Maquet,
1961). Bloom et al. (1986) argue that for less developed
countries facing the problem of overpopulation, the greatest
challenge is "to generate sufficient jobs at reasonable
wages to absorb their rapidly growing populations into
productive employment." The dilemma, they suggest, is that
population growth has an impeding effect on wages and
employment.

In rural areas of Rwanda, where in some parts the
population has reached 1000 persons per square kilometer
(Habyalimana, 1990), it goes without saying that land has
become a scarce resource. Not only is production weak, but
the creation of jobs at reasonable wages is made doubly
difficult by the urban bias of most industries. Consequently
there is a significant income differential between rural and

1

2
urban areas. The increasing scarcity of land in the rural
areas is likely to have a negative effect on the
distribution of landholding in Rwanda, an effect that will
widen the gap between large and small land holders.

The combined problems of overpopulation, slow
agricultural growth and below-subsistence incomes earned by
major segments of the farm population, have compelled many
to seek employment in the non-farm sector. It is postulated
by Clay, et al. (1989: p.1) that income from non-farm
employment will help alleviate the hardships that ensue from
Rwanda's rapid population growth and low production, and may
reduce the income gap observed between rural and urban
areas and among segments of the rural population itself.

Off—farm employment has long been recognized by

farm populations as a mechanism for bridging the

income gap created by stagnant farm production and

ever-increasing population pressure... Beyond its

contribution to the overall growth of rural

incomes, expansion of the non-farm sector has the

potential for alleviating income inequalities

inherent in the agricultural sector, notably those

associated with an unequal distribution of

landholding.

In the same way, it is believed that off-farm labor
helps to improve the status of the poor. This has certainly
been the case in Asia but, as Hazel et al. (1987) caution,
this conclusion cannot be imported wholesale to the
countries of Africa. However, some activities, notably food
preparation and others dominated by women, do tend to reduce

differences in income. They further state that "even if the

impact on relative inequality is uncertain or negative, in

3

terms of absolute poverty alleviation non-farm earnings
cannot but help improve the economic status of the poor."

Based on data from a recent survey of 1,099 farm
households, this study examines the structure of non-farm
employment in Rwanda. Special attention is given to the
situational circumstances under which households look to
such employment as a matter of necessity and/or as an
alternative to traditional work roles. In other words, to
what extent are younger farmers pushed out of agriculture
because of insufficient landholding or, by contrast, does
the non-farm sector select out only those with the
educational skills necessary to compete, regardless of their
resource endowment. Additionally, this study explores the
importance of non-farm employment opportunities vis-a-vis

the inequalities in land and income distribution in Rwanda.

NON-EARN EMPLOYMENT: CONCEPTS AND LITERATURE

Research on the emergence of non-farm employment in the
Third World has been the focus of attention of many social
scientists in recent decades. Based on an analysis of data
from International Labor Office (ILO), Carl Liedholm (1973)
concludes that rural non-farm activities comprise a
significant part of all rural employment. In Nigeria, for
example, ILO data show that 41% of employed males were
engaged in rural non-farm activities in 1970. In Uganda,
estimates were in the range of 20%. Elsewhere Liedholm

(1973) determined that the level of non-farm activity varies

4
directly with population size, and that such employment can
be seen as a partial solution to the problem of growing
population pressure. Though pervasive, the extent of non-
farm activities is often difficult to assess accurately
because they tend to be found in the informal sector and in
small and micro enterprises. It is reported that in Africa
most rural non-farm employment is generated by micro-
enterprises and firms with fewer than 10 workers (Liedholm
and Mead, 1987).

More recently, Liedholm and Mead (1987) have
underscored the role of small-scale industry in generating
employment. They observe that small-scale industry is a
reaction to the failure of modern industrialization
strategies to solve the problem of unemployment and poverty.
By comparing six developing countries that cover a broad
range of wealth and poverty as measured by per capita GDP in
1980 (Jamaica, Thailand, Honduras, Egypt, Sierra Leone,
Bangladesh), they discovered that some of these countries,
Bangladesh for example, had very low per capita GDP yet
relatively high numbers of small enterprises. They also
examined growth rates and found that some, like Sierra Leone
and Jamaica, were actually experiencing negative growth.
Although the general trend was an overall increase in
manufacturing output during the 19708, the authors found
that "employment in large scale manufacturing firms did not
keep pace in any of these countries" and was even declining.

Data from six other countries (Haiti, Zambia, Burkina

5
Faso, Botswana, Indonesia and Kenya) have been cited in
support of their argument. Evidence drawn from an empirical
study of 14 other countries showed that small scale
industry, firms employing less than 10 persons, account for
more than 50% of total industrial employment.

Because of the linkages between small scale enterprises
and low per capita income, Liedholm and Mead (1979) conclude
that such enterprises tend to emerge in lower income
countries. The lower the income per capita, the higher the
percentage of employment generated by small scale
industries. Although providing small relative value added,
small scale industries have a substantial effect on the
economy. Small-scale industries are most often engaged in
the production of "light" consumer goods, principally food
and beverages, textiles and wearing apparel, and wood
products. In a ten country survey the authors also found
that such small-scale industries were primarily located in
rural areas for most of the countries and were generating
more employment than the more heavily industrialized urban
firms. In a sample of 13 developing countries, it was found
that 63% of the manufacturing employment was generated in
rural areas. Because of the informal, less visible nature
of these rural enterprises, some suspect that the percentage
might be higher still (Liedholm and Mead 1979).

Other important questions deal with the source of
capital used by small industries, their rates of growth, and

employment levels. Looking first at sources of capital, it

6

is reported that most of the funds invested in small-scale
industries originate from personal savings or from friends,
thus showing the limited involvement of the Government and
other formal credit institutions (Liedholm and Mead, 1987).
Data from five countries (Bangladesh, Nigeria, Sierra Leone,
Tanzania and Haiti) were examined in support of this
argument. Concerning the growth rate, it was found to be
generally quite low among small-scale industries in India,
Ghana, Sierra Leone, Turkey and Columbia, although its
overall contribution to employment is substantial. Steel
(1977) reports for example that during the 1960's smaller
firms did absorb 6 times as many new manufacturing workers
as did the large scale industries. Hoselitz (1959) and
Anderson (1982) identify three phases in the process of
industrialization. Based on these three phases, Liedholm
and Mead (1987) compared changes in the size of small-scale
industry to changes in the level of industrialization. In
phase I or the early stage, they found that household
activities dominate; in phase II, small and medium scale
activities emerge and some household activities are
displaced; in phase III, large scale industry predominates.

Another consideration is the demand for goods produced
by small-scale industries. Although small-scale industry
helps to boost rural household incomes, some studies have
argued that the demand for goods from small-scale industries
in rural areas will decrease as household incomes increase,

because such goods are often of inferior quality (Resnik and

7
Hymer, 1969). By contrast, others argue that increasing
rural incomes will raise the demand for goods produced in
rural areas (ILO Mission 1978, Mellor 1976, Chuta and
Liedholm 1979). Empirical evidence from Nigeria, Malaysia
(Hazel and Roell, 1983), Sierra Leone (Byrlee and King,
1978) and Bangladesh (Deb and Houssain, 1984) demonstrates a
strong positive correlation between change in rural
household income and changes in the demand for rural small-
scale produced goods.

The role of the State in promoting smalléscale industry
has been examined by Liedholm and Mead (1979) who point out
that the State governments have not been active buyers of
small-scale industrial goods. There are many ways in which
Governments can enhance or hinder the development of small-
scale enterprises, but particularly through the regulation
of raw material imports. These imports may be directly
allocated by centralized agencies or state trading firms,
which often disadvantage small producers. Such was the case
in Egypt where small producers encountered a severe problem
until the government established an agency for channelling
imported inputs to small producers in cooperatives.

Another example of government intervention is the
creation of tariff structures that discriminate against
certain types of enterprise. It is reported that in Sierra
Leone, for example, "textile machinery used by large
garment manufacturers was imported at a zero tariff rate.

Yet simple sewing machines, basic items of capital for small

8
tailors were subjected to a high duty on the argument that
they were considered luxury consumer goods." (Chuta and
Liedholm, 1985)

Governments also influence the market for goods
produced by small-scale industries. In Korea, for example,
government regulations all but eliminated the incentive for
exporting and importing goods from small-enterprises. Frank
et al. (1978) show that minimum export and import values
were set such that small firms were not eligible for
subsidies. They state that,

To maintain a privileged status, traders had to

sustain annual exports exceeding $20,000 per year

for firms registered as exporters and $100,000 for

importers. Many small enterprises were unable to

meet these volumes. In Taiwan on the other hand,

where the policy stance has been less

discriminating, small firms have played a much

larger role in industrial development as well as

in exporting. (Haggblade et al., 1986).

In short, governments can play an important role in
developing the rural non-farm sector, notably through the
regulation of imports and exports, tariff structures and the
availability of credit. But from a supply side perspective
the willingness of governments to favor small enterprise
development will and should depend at least in part, on the
relative efficiency of small enterprises when compared to
the heavier industries of the formal economy.

Liedholm and Head (1987) argue that governments should

selectively channel economic resources to small-scale

industries that are socially more efficient than larger

9
scale industries and affirm that this could be beneficial in
creating output and employment.

In their review of literature they conclude that, in
general, small-scale enterprises in developing economies are
efficient in the use of resources. They add that because of
their contributions to income and productive employment, it
is appropriate that they be the target of policy and
project-focused attention from Governments as well as from
international donors. Despite this observation, it is also
found that "in most countries of the Third World, the
overall policy environment is skewed against small
producers. This is primarily because policy is aimed at
providing special benefits to larger manufacturing
enterprises." (Liedholm and Mead, 1987).

It is also argued that there is a linkage between
agricultural and rural industrial activities and that
agriculture generates the largest share of rural income.
Because of this, any policy aimed at increasing agricultural
output has an indirect effect on the demand for products of
rural small-scale enterprises. Increasing agriculture
output will increase rural incomes and the demand for rural
small-scale industrial goods and services. This is
especially so for small farmers who make up the largest
segment of the rural population in the developing world.
Goods produced by the industrial sector are more likely to

be found on larger, commercial farms in these areas.

10

Assistance to small-scale industries is crucial in many
different areas: in the provision of credit, technical]
production and assistance, management assistance, marketing
assistance and common facilities, just to name a few. In
the area of credit, we know that a shortage of working
capital is a major constraint faced by small-scale
industries in many countries including Sierra Leone (Chuta
and Liedholm, 1985), Haiti (Haggblade et al., 1979),
Bangladesh (Bids, 1981), Honduras (Stallman, 1983), Thailand
(Narong Chai et al., 1983) and Jamaica (Fisseha and Davis,
1981). Despite awareness of the credit problem, it is argued
that "financial institutions have been reluctant to expand
their lending to small-scale firms."

Although each country is in some way unique and must be
considered individually, there are commonalities which
permit us to generalize in our claim that small and micro
enterprises are essential to the development of Third World
agricultural economies. This position is strongly supported
by Harper et al. (1979) who assert that "the attempt to
develop indigenous small enterprises is as legitimate and
important as a function of economic development as other
more traditional tasks which have received official
attention for a longer period."

Similarly, because of the contribution of small-scale
enterprises in employment and income improvement in
developing countries, Liedholm and Mead (1987) conclude that

Governments in developing countries must seek changes in the

11
general policy environment that broadly affects small
private enterprises, and implement specific projects

designed to provide direct assistance to individual firms.

Non-farm Employment in the Rwandan Context

Many of the observations made and conclusions drawn
above are applicable to the Rwandan case. Ecological theory
tells us that a more differentiated employment structure
(i.e., specialization in the non-farm sector) constitutes
one possible solution to the problem of population pressure
on the land (Durkheim, 1933). Other alternatives include
the intensification of agriculture (Boserup, 1965, 1981),
territorial expansion or migration (McKenzie, 1933; Hawley,
1950) and reduction of human fertility (Bilsborrcw, 1987).
In Rwanda, public and private initiatives have been
developed to increase agricultural production through
intensification, to facilitate the movement of households to
other regions (e.g., the paysannat resettlement program) and
to bring down Rwanda's fertility rate through family
planning, but relatively little has been done to stimulate
growth in the rural non-farm sector.

It is conceptually useful to distinguish between non-
farm employment which implies the sector of employment
activities and off-farm employment which implies the
location of activities, both of which, together, comprise

the process of occupational specialization. These criteria

12
have been discussed in the Rwandan context by Mead (1989)

and have been summarized as follows.

MOW

 

Sector of

 

 

Activity On Farm Offr1§g1L______
Farm I Agriculture II Agric. Labor
Non-farm III Non-Farm Activ. IV Off-Farm Activ.

 

Source: Mead (1989)

This table illustrates the intersection of these two
criteria, and the four resulting categories of employment.
Zone I contains agricultural activities performed on one’s
own farm. Agricultural activities performed away from the
individual’s own farm, typically as an agricultural laborer,
can be found in zone II. Zone III represents activities
performed on the farm but not directly related to
agriculture; fabrication of pots and clothing are examples
of such activities. Finally, zone IV categorizes work
performed off the farm and not in the agricultural sector,
including retail sales, carpentry, masonery and others in
the building industry, to name just a few. The focus of
attention from this point onward is on "non-farm employment"
(either on farm or off farm, i.e. category III and IV) and
on "agricultural labor" (category II), since together these
activities constitute the principal employment alternatives
to subsistence agriculture for Rwandan farm households.
Subsequently, non-farm employment and agricultural labor

will be used to mean both non and off-farm labor.

13
Structure of Employment

Agriculture forms the backbone of the Rwandan economy,
as 93% of the population live and work in rural areas
(Bureau de recensement, 1978). Yet because economic growth
has a strong urban bias, Rwanda’s rural areas have become
the principal supplier of labor for urban areas. This rural
exodus has gained great momentum in recent decades (Olson,
1990), creating a flow of migration that may never be
absorbed by the urban economy. Rising urban unemployment,
high crime rates, poor housing and sanitation, and the
potential for civil unrest are but a few of the symptoms of
this transformation as it has been experienced around the
world. Historical evidence shows us how "push" factors,
notably scarce and unequally distributed land resources, the
lack of alternative employment opportunities, and a low
standard of living, invariably contribute to rural
depopulation and to urban saturation. Improving
opportunities in the rural sector and eliminating the urban
bias may be the best hope for many developing countries,
including Rwanda, to confront rural poverty. Byerlee and
Eicher (1972) assert that since about two-thirds of the
population in most African countries live in rural areas,
national policies to deal with the employment problem will
depend to a large degree on the ability to develop
appropriate strategies and policies for rural development.

A recent study found that more than 50% of employment

in manufacturing was engaged in the brewing of traditional

l4
beverages. Eighty percent of the remainder was engaged in
un-registered activities such as brick and tile makers,
tailors, basket makers, etc., and over 75% of this
employment was in rural areas. As discussed above,
insufficient landholdings have compelled farmers to split
their time into farming and non-farming activities, hence
some of these activities are performed on a part-time basis.
In the Rwandan context, this notion has special meaning.
Part-time and non-farm activities are often intertwined for
those operating small farms (Barlett, 1986). Small land
owners, generally the poorest in rural areas, are invariably
the most likely of all to turn to the non-farm sector in an
effort to supplement earnings from agriculture (Goldkind,
1961). Though less time is devoted to agriculture in this
instance, earnings from non-farm sources may be used to
improve agricultural production. Improvements in
agriculture will in turn boost rural incomes. In this way,
non-farm activities often go hand in hand with improvement
in agricultural production. However, not all non-farm
earnings are reinvested in the farm enterprise. More often
they are used to meet basic consumption needs that have not
been satisfied by earnings from the farm.

The net impact of such changes on the rural poor is
conditioned by the specific form of employment opportunities
available. In areas where agricultural labor is the
principal employment alternative for those with insufficient

landholdings, the emergence of groups based on landholdings

15
often occurs. Many, beginning with the classical theorists,
have observed the bifurcation of the farm population into
“landholders" and wage laborers (Marx, 1967). Those with
holdings insufficient for their families' subsistence needs
are pressured to seek gainful employment off the family
farm. In turn, those possessing excess holdings can put
this surplus labor to work for a set wage.

Where initially these two groups are unified, as
peasant farmers concerned with producing for their own
subsistence, resource scarcities force their separation.

The farm manager is oriented toward production for a surplus
and enters into a competitive market where this surplus is
exchanged for a profit. The agricultural laborer also loses
his identity as a peasant farmer and competes in a
competitive market for the sale of his labor (Mendras,
1970). By converting surplus production into increased
landholdings, and/or higher levels of technology, the two
groups become further differentiated and, to the extent that
their holdings and occupational roles are maintained across
generations, this process of differentiation and
specialization leads to the formation of two identifiable
social groups.

Though differentiation in the agricultural sector tends
to be closely tied to landholdings, differentiation in the
rural non-farm sector can occur independent of the
distribution of landholdings, and benefit even those whose

on-farm production capabilities are limited. By creating

16
employment opportunities for the poor and near-landless
segments of agrarian society, growth of the rural non—farm
sector can effectively begin to alleviate some of the
inequalities inherent in the agricultural sector.

In summary, though the growth of non-farm employment
opportunities is vital to rural development effort in the
Third World, the consequences of such growth for
inequalities and in the distribution of incomes are
potentially very strong. This paper examines patterns of
rural non-farm employment in Rwanda and the extent to which

such activities are linked to land and income inequality.

an-ferm Employment and Income Distribution
In almost all countries that engage in non-farm activities,
the trend is such that those activities have a great impact
on income distribution. The effect of non-farm employment
on income inequalities is variable. In certain contexts
non-farm earnings tend to favor wealthier households and
hence widen income inequalies. In other cases, the poorer
households tend to be the primary beneficiaries. Haggblade
et a1 (1987) share with us examples of where non-farm
earnings, despite their importance in raising overall levels
of income, do not bring about greater equality. Studies in
Nigeria, Lesotho, Tanzania, Zambia, Uganda and Zimbabwe have
shown that non-farm income is linked with higher income
groups, though all groups benefit in absolute terms. Other

studies, notably those condected in Zambia, Botswana and

17

Nigeria can be cited as examples where non-farm employment
tends to benefit poor households more so than the wealthy.
These cases tend to emphasize the importance of women’s non-
farm earnings in reducing inequalities. Although these
seemingly contradictory studies do not permit us to draw any‘
general conclusions about the role of non-farm employment in
promoting the status of poorer households, they do provide a
formidable challenge to those interested in exploring this

relationship.

HISTORICAL AND SOCIO-CULTURAL CONTEXT

Located in the highlands of East Africa, Rwanda is a
country of mountains, hills and great lakes. This
landlocked country shares its northern border with Uganda,
while former sister-state Burundi is located immediately to
the south, and Tanzania to the east. Rwanda is also flanked
by Lake Kivu and a volcanic mountain range which occupies
much of its border with Zaire to the north and west (Figure
1).

The history of land settlement in Rwanda has been
passed down through the generations by oral tradition.
sifting through twentieth century literature reporting the
works of the chroniclers on the "Sacred Kingdom," one learns
of the settlement patterns of the various ethnic groups that
make up the country' 3 population today. The first group to
occupy the territory now known as Rwanda were the "Twa." The

Twa had little interest in agriculture and derived

18

 

   

_z_.

     
 

w h
Q’
\
Y
N
a Kibungo
“(nonzero
Cyangugu

Butare

 
    

 

BURUND‘
0 20 40km
L.L__LI_J

 

 

 

 

 

 

Figure 1. Map of Rwanda (source: Olson, 1990)

sustenance through hunting and gathering activities. The
exact moment of their arrival is uncertain due to the
incompleteness of historical records. The Twa were later
joined by the "Hutu," agriculturalists who are believed to
have originated from the west as part of the Bantu
migration. A third ethnic group, the "Tutsi," is reported

to have made their way to the highland areas in the 14th

19
century (Kagame, 1972). Arriving with herds of cattle, the
Tutsi adapted slowly and peacefully to local conditions and,
claiming to have come from divine origin, eventually wrested
rights to the land from the Hutu majority, and ruled under a
feudal system called "ubuhake" (to submit to the will of
others).

Through the ubuhake system as reported by written
history, the Hutu obtained the use of cattle owned by the
Tutsi rulers and also received protection from this "would
be elite group." In return, the Hutu were obliged to work
for the Tutsi, to surrender a part of their harvest and to
serve in the king's army as rank and file soldiers (Nyrop et
al, 1985). This system was maintained through the
centuries, and prevailed even during periods of German
colonization and Belgian "mandate and tutela” colonial rule.
In 1959, however, the internal social revolution that led to
independence in 1962 successfully dismantled the traditional
feudal structure and created a more equitable system of
independent peasant landholders.

The total population of Rwanda was estimated at
3,073,000 million in 1967 (Rwanda, Ministry of International
Cooperation and Plan, 1967) and the average density of this
population was more than 163 persons per square kilometer.
Current figures show that due to rapid population growth,
the size of Rwanda's population is now in excess of seven

million persons. At an annual rate of growth of 3.7 %,

20
Rwanda’s population is expected to double in less than 20
years (Population Reference Bureau, 1989).

Hilly and sloping from west to east, the average
altitude is approximately 1,600 meters with the mountainous
western provinces having the highest altitude. The central
and eastern provinces are lower in altitude, yet still hilly
topographically. The capital city, Kigali, is located in
the central eastern part of the country. There are four
seasons, each with its corresponding agricultural
activities. Two rainy seasons (September-December and
February-May) are associated with agricultural seasons, and
two dry seasons match the time of harvest and the clearing
of the land for the next season.

The economy is based on agriculture, mostly subsistence
agriculture. Beyond meeting the subsistence needs of their
own households, most small farmers also produce cash crops,
notably coffee and tea. Subsistence and cash crops compete
on the small holdings (averaging 1.21 ha) and almost every
household is involved in both types of production.

The use of modern technology in subsistence crop
production was unknown until almost two decades ago. Today,
one aspect of that technology, the use of organic
fertilizers, is the most common among all the farmers who
rely on it for increasing soil fertility. These fertilizers
are produced locally at the individual household level.

Virtually all agricultural tasks are accomplished with

the aid of simple hand implements. Hoes are used to till

21
and weed the soil, and machetes to clear the land and
harvest certain crops such as sorghum and bananas. Neither
animal traction nor motorized equipment are used.

Among the most important crops are bananas and sorghum
which are grown primarily for use in the production of beer.
Even the most casual social gatherings are marked by a
sharing of conversation and the consumption of home-brewed
beer, either banana or sorghum or both. Other important
crops are beans, maize, potatoes, sweet potatoes, cassava,
peas, peanuts, rice, and fresh vegetables.

Population pressure and declining soil fertility have
prompted farm households to use marginal lands to produce
rice and cassava. Rice and potatoes are the sole crops
produced using modern technologies and there is no major
food processing industry in Rwanda, though some small-scale
commercial processing of dairy products such as milk, cheese
and yogurt has developed in recent years to serve the urban
market (Rwamasirabo, 1990). In the same way, but oriented
more toward rural areas, a maize processing plant for flour
production has been put into operation.

The traditional division of labor is based on gender
and age and reflects the historical domination of women by
men. Women split their time between agricultural and home
based activities. In the field women take on the major
responsibility for sowing, weeding, harvesting and
processing crops after harvest. Cooking, cleaning and child

care define life for women in the home. Young children are

22
expected to help their mothers both in the fields and at
home, particularly with cooking, collecting firewood,
fetching water, etc. Older children follow in the footsteps
of their parents according to gender. Girls prepare
themselves for motherhood and boys for fatherhood. Men are
responsible for clearing land and share in its tilling, but
their involvement in the fields beyond that point is limited
to the maintenance and harvesting of coffee, bananas and

other cash crops.

INMEA AJHDIKETTKN)

Data analyzed in this paper were obtained through a
nationwide random sample of 1,0191 farm households in Rwanda
as part of the 1988 Non-farm Strategies Survey. Survey
questionnaires were administered over a three-month period
beginning in July, 1988, to various members of sampled
households, including husbands, wives and adult children.

An experienced team of Rwanda's Agricultural Statistics
Division (DSA) field staff supervisors was engaged to carry

out the interviews.

 

1An unusual yet important feature of this study is that the
households selected for study are actually a sample of households that DSA.
has been studying since 1986 as a part of its ongoing survey program.
Consequently, the great volume of information already collected on these
households, in areas such as farm size and fragmentation, crop and
livestock production, and market transactions, can be drawn upon to
enhance our analyses of non-farm strategies. Since 1986, a small number
of households in the initial sample have moved away or have been otherwise
dropped from DSA's current sample. The total number of households on
which data are available at all points in time is 1019; these are the
households for which data are presented in the following section.

23

The questionnaires were designed to obtain information
from various members of the households, including husbands,
wives and adult children. Topics addressed in the
questionnaires included: demographic characteristics of all
household members and migrant children; non-farm and
off-farm employment; permanent and temporary migration
patterns of selected household members; fertility/family
size behaviors, plans and preferences of all adult household
members; economic support networks between the household and
members of the extended family living elsewhere; sources of
household income; physical characteristics of the farm and
residence; hired farm labor; and the plans, aspirations and
opinions of parents and adult children regarding non-farm
training and employment and the future for young people in
farming.

In all, interviews averaged approximately one and a
half hours to administer and usually required multiple
visits to meet up with the various respondents from each
household. Adult children were the most difficult of all to
locate, though through multiple call-backs the interviewers
did succeed in interviewing a large number of these young

people.

FINDINGS AND DISCUSSION
Patterns of Non-farm Employment
Approximately half (47%) of farm households in Rwanda

engage in some form of non-farm employment, and since our

24
survey reference period includes only the three months
preceding the date of the interview, it is very likely that
this figure underestimates the true proportion. When the
members of farm households do seek non-farm employment, it
is most commonly as agricultural wage laborers (31%) on the
farms of their neighbors (Table 1). The remaining 69% of
their time off-farm is spent in the rural non-farm sector,
notably as artisans, laborers or in commerce -- generally in
small businesses. Over five percent of non-farm employment

is held by government functionaries.

Table 1. Percentage of Days WOrked in the Non-farm Sector
or in Agricultural Labor

T o m lo ent % of a s
Agriculture labor 30,8
Artisanal 18,9
Non-farm unskilled lab. 12,8
Commerce 16,4
Functionaires 5,6
Other salaried 12,3
All others 3,1

Total 100,0
(N=) (454)

Those engaged in agricultural wage labor are most often
hired to clear and till the soil or to weed and otherwise
maintain fields of crops. Only 6.3% of all hired
agricultural labor is used at the time of harvest. The
three most important crops for which labor is hired are
beans, sorghum and sweet potatoes (DSA, 1987). Time spent
off the farm in artisanal activities is heaviest in the

construction industry. Brick and tile makers, masons and

25
carpenters account for over half of all artisanal
employment. Basket weaving, tailoring and art/embroidery,
all industries in which women are heavily involved, comprise
another 37% of the artisanal subsector.

The number of days per year that households are
employed off the farm in various types of activities is
presented by farm size category in Table 2. The estimated
average number of days of non-farm work by Rwandan
households is 78, or slightly under a third of a person-

year.2 As described above, agricultural wage labor,

artisanal trades and commerce constitute the largest portion
of this time. Households in the smallest farm size category
appear to be the most likely of all to seek employment off
the farm. This finding is consistent with our initial
hypothesis that households experiencing the greatest
resource constraints (insufficient landholdings) would be
more actively engaged in non-farm employment.

The only farms to depart from this otherwise very
supportive pattern of findings are those in the largest farm
size category (2.0 Ha and above), which work off the farm a
relatively high 77 days per year. An investigation into why
these households might be employed in non-farm activities
more than expected, found the explanation to be that these
large farms tend also to support relatively large

households, and the more adult workers there are in a

 

2 Estimate based on 250 work days per year.

26
household the higher the chance that one or more of these
workers will be "freed up" for non farm work (Clay et al.,
1989). By controlling for the number of adult workers in
the household we find that non-farm employment among these

households drops into the expected range.
—

Table 2. Days of Non-fare Work by Type of Employment and Fare Size

 

 

Umumjmmnmmt

Fare Size Ang- Arti- Unskld Come- Other
(he) _!_ on t l r r F ther Total
<0.5 253 31.2 16.8 10.4 10.8 4.0 20.4 3.2 96.8
0.5 - 1.0 325 30.8 14.8 9.6 8.8 0.4 7.6 3.6 76.0
1.0 - 2.0 261 15.6 14.4 10.0 16.4 1.6 5.6 0.2 63.6
2.0 + 180 12.8 14.0 10.0 16.4 17.2 4.8 1.2 76.8
Total 1,019 24.0 14.8 10.0 12.8 4.4 9.6 2.4 78.0
Sign. .003 .093 .996 .325 .001 .056 .317 .050

The type of non-farm employment households engage in is
also connected to the size of their holdings. Households
with small landholdings tend to work off their farms as
agricultural wage laborers, while those with larger holdings
are more likely to hold jobs as functionaries and in
commerce, though differences in the latter are not
statistically significant. Employment in the artisanal
trades seems to bear little connection to farm size. In
summarizing the pattern of findings reported in Table 2, one
can conclude, at least tentatively, that those from large
farms tend to occupy those jobs that require higher levels
of schooling and working capital, while those from
households with small holdings make up the bulk of the

agricultural wage labor pool.

27
Non-farm Employment and Characteristics of Individuals

To learn more about how and why certain households look
to non-farm employment in an effort to generate income, we
examine some of the characteristics of the individuals who
comprise these households and who engage in non-farm
employment. When households do seek such employment, 77.5%
of the time they are represented by only one person. Less
frequently, a second household member (15.9%), or even three
or more (6.6%) can be so employed.

Table 3 reports that heads of households work in non-
farm sector far more often than do other household members,
as they account for over 55% of all non-farm employment in
Rwanda. Adult children still living in the household are
next at 34.3%, followed by spouses at 7%. Consistent with
these results is the finding that nearly 80% of all non-
farm employment is held by men. Because traditional Rwandan
society places a heavy burden on women to labor both in the
fields and in the home, few women are encouraged to seek
specialized vocational training. The areas where women make
the largest contribution off their own farms are as
agricultural wage laborers, basket weavers and seamstresses
-- all jobs that have great flexibility as to either when or
where they are performed, thereby permitting coordination
between farm and non-farm responsibilities.

Age also appears to be an important determinant of non-
farm employment. Roughly 60% of this labor market is

controlled by workers in their 20s and 303. Those in the

28

Table 3. Percent of Non-farm Employment and Mean Days Worked per
Year by Characteristics of Individuals (Population 8 all
individuals who work in non-farm sector).

 

% of Mean Days Distribution
Selected Non-farm Worked of Population
acts 3 ice 0 t Peg Yea; Age 10+ Years
Relation to head
of household:
Head 55.7 134 20.0
Spouse 7.0 78 14.8
Child 34.3 124 60.6
Other 3.0 66 4.6
Total 100.0% 124 100.0%
Sex:
Male 79.3 140 47.0
Female 20.7 85 53.0
Total 100.0% 124 100.0%
Age:
10 - 19 13.5 111 29.9
20 - 29 31.2 140 25.2
40+ 25.0 120 22.8
Total 100.0% 124 100.0%
Education level:
None 35.6 106 41.8
Some primary 33.2 111 40.8
Primary + 31.2 177 17.4
Total 100.0% 124 100.0%
(N8) (667) (667) (5,084)

20-29 age bracket invest a greater share of their time in
non-farm jobs than any other group, averaging 140 days per

year. There are several good reasons for why this age group

29
stands out in its provision of non-farm labor. The first is
that these young people are still experimenting a great deal
with alternative career strategies and, because they have
fewer dependents (or may still be living on their parents'
farms), stability of employment is less crucial. Second,
parents of these young people are in their 40s and 505 and
are not yet ready to pass on more than fragments of their
land holdings to their children. Third, many of these young
farmers are aware that they will never inherit enough land
to meet the subsistence needs of their families and believe
they must make a niche for themselves elsewhere. Fourth,
the level of schooling and specialized vocational skills
obtained by this group exceeds that of their elders,
allowing them to compete for those jobs in the non-farm
sector requiring higher levels of training.

Table 3 highlights the importance of educational
attainment to securing non-farm employment, as nearly a
third of all such employment is held by the 17.4% of
individuals who have managed to complete primary school.
Those who comprise this "elite" group perform non-farm work
an average of 177 days out of an estimated annual 250 work
days, or 71% of their time. By comparison, individuals with
no formal schooling at all work in non farm sector only 42%
of the time. Globally, those who work in non-farm work do

so 124 days per year on average, or "half time."

30

Multivariate analysis of Agricultural Wage
Labor and Non-farm Employment

In previous sections we examined some of the bivariate
relationships between characteristics of households and
individuals on the one hand, and non-farm employment on the
other. We now turn to a multivariate regression analysis
that will enable us to assess the relative importance of
some of these variables in explaining variations in non-
farm employment and the sale and purchase of agricultural
wage labor.

Table 4 presents findings from three independent
multiple regressions. The first two columns report the
simple correlation coefficient (Pearson’s r) and the
standardized regression coefficients (beta), between the
number of days households worked in non-farm activities and
seven household characteristics (predictor variables). The
simple correlation coefficients indicate the degree of
association between two given variables, without controlling
for the effects of other explanatory variables, while the
beta coefficients represent the correlation between two
variables once all other identified explanatory variables
are held constant (i.e., the true independent effect).

The strongest predictor of household participation in
the rural non-farm sector is the number of adult male
workers in the household (beta=.26). A large number of
workers can easily exceed the labor requirement of the farm.
Rather than under-employing these hands within the

household, they are encouraged to seek alternative

31

Table 4. Multiple Regression Analysis of Days of Non-farm
Activities by Household Members in the Non-
agricultural Sector and in the Agricultural Sector,
and Days of Wage Labor Hired by Households.

Days non-ag Days ag labor Days ag labor
WM—w

 

 

Explanatory variables beta (r1, beta (:1 that; (r)
Days non-ag employment -- -- -.12* -.07* .17* .19*
Days Ag labor worked -.12* -.07* -- -- -.02 -.07
Days ag wage labor hired .17* .19* -.02 -.07 -- --
Farm size -.11* .04 -.15* -.10* .19* .20*
Male workers in hshld .26* .25* .21* .15* —.07 .03
Female workers in hshld .03 .07 .05 .06 -.01 .03
Educ of head of hshld .12* .17* -.01 -.02 .16* .19*
Age of Head of hshld -.01 -.05 -.06 -.08* .05 .03

Multiple R .35 .25 .32

* Significant - <.01

e------------------------------
employment opportunities in non farm activities. By
contrast, an increased number of female workers does not
translate into increased non-farm employment. This finding
reflects the fact that women generally lack the training
necessary for non-farm employment and, concomitantly,
because they are assigned so many other household
responsibilities such as caring for young children and
preparing meals, their absence from the household tends to
disrupt established patterns of daily life (and male
domination) far more than does the absence of male workers.
Education of the head of household similarly has a
positive effect on household employment. As suggested
earlier, this is largely a reflection of the fact that non-

farm employment so often requires higher levels of formal

32
schooling. This is true for the independent trades and
small businesses, as well as for those who are salaried by
local enterprises or the government.

The zero-order correlation between farm size and non-
farm employment (r=.04) is slightly positive in direction
but not statistically significant. This initial finding
contradicts our theoretical argument that households with
the smallest landholdings are forced to resort to employment
in the non-farm sector. However, once we control for other
variables in the model, the hypothesized negative
correlation between farm size and non-farm employment
emerges (beta=-.11). In other words, because large farms
tend to have many adult workers and their heads of
households tend to have achieved relatively high levels of
schooling, two characteristics that have also been shown to
lead to non-farm employment, the correlation between farm
size and non-farm employment is "forced" in a positive
direction. However, once the effects of these other
variables are held constant, the true influence of farm size
on non-farm employment becomes visible.

We now turn to the set of household characteristics
that induce members of households to seek outside employment
as agricultural wage laborers. In Table 4 we note, first of
all, that households whose members seek employment as farm
laborers are not, generally speaking, the same households as
those whose members find employment in the non-farm sector

(beta=-.12). Despite the tendency to work in either one or

33
the other sector, these two groups of households have two
things in common -- they both tend to have relatively small
landholdings and a large number of adult male workers.
Again, this finding supports the notion that non-farm
employment serves as an important strategy for households
suffering from an over supply of labor relative to
landholdings.

The third set of regression coefficients in Table 4
pertain to the amount of farm labor that households hire in.
It is no surprise to find that, unlike households that
engage in large amounts of non-farm labor, households that
hire in labor tend to operate larger farms (beta=.19) and to
have relatively few adult male members (beta=-.07), though
this latter relationship is not statistically significant.

Perhaps most revealing of all is the finding that
households that hire the most agricultural labor also tend
to work in the non-farm sector (beta=.17), notably in
commerce, in various trades, and as functionaries. The
higher education levels attained by these farmers permits
them to participate in the higher end of the non-farm
employment market, and pay minimum wage rates to those who

work their holdings.

Focus on High Income Households
To further explore the notion that farm size, education
level, non-farm employment and the employment of

agricultural wage laborers are four variables that tend to

34
mutually reinforce one another in the process of income
generation, and implicitly, of income inequality, we have
isolated the 51 (5%) highest income households in our
sample. These households generate an average income of
253,000 Frw per year, approximately 3.6 times the average
for all households outside of this high income group.

As expected, we find that many of the heads of these 51
households have achieved relatively high levels of formal
schooling. More than a third have completed primary school,
and many of these have gone on to even higher levels. Due
in large measure to the high educational attainment these
farmers, and other members of their households, their
participation in the non-farm sector is exceptionally high.
Over two-thirds of these households generate non-farm income
from one source or another, and of these, 77% are employed
as functionaries or small businessmen (the upper echelons of
the non-farm sector).

Though these high-income households are heavily engaged
in the non-farm sector, they also operate (manage) some of
the largest farms in our broader study sample. A full third
of these households operate holdings of 3 hectares or more,
and the average for this group of 51 households is 2.6
hectares. Generally speaking, these should not be
considered subsistence-oriented farms. Rather, they are
market-oriented and make heavy use of hired farm labor. Of

the 51 farms, 37 (73%) employ labor from outside of the

35
household, and four of these employ the equivalent of three
or more full-time workers.

This brief analysis of high-income households serves to
highlight the combined importance of education and
landholdings as the basis for income generation in Rwanda.
By applying these human and land resources to an
increasingly differentiated occupational structure, these
households can draw from both the high end of the non-farm
sector and from a growing pool of agricultural wage labor,

to further expand their income generating capabilities.

Non-farm Labor and Inequality

The polarization of farms in Rwanda in terms of their
ability to generate income can be seen in Table 5 which
shows the proportion of farms in various income categories
and the proportion of total income (column 3) they generate.
While the households in the highest income class comprise
only 15% of the population, they command nearly 35% of the
country's total rural income. This concentration of wealth
is represented in a Gini coefficient of .34, which is
comparable to many other African countries.

For purposes of the present analysis, however, our
interest is more to assess the extent to which non-farm
employment helps reduce income inequality. Where column 3
of Table 5 presents the distribution of total household
income in Rwanda, columns 4 and 5 break out this total into

sub-categories -- "agricultural" and "non-agricultural"

36

Table 5. Percent of Households, Total Income, On-farm and Non-
farm by Income Group

 

 

% of % of % of

Income group % of Total Agric. Non-agri.
(in FRW) 4ggg§§hgld§ Income, Income Income
Less than 33,000 15.2 4.5 4.9 1.3
33,001 - 50,000 18.6 9.5 10.4 3.9
70,001 - 90,000 15.6 15.4 15.6 13.8
90,001 - 125,000 15.9 21.0 21.3 19.1
125,000 and above 15.2 34.9 32.8 49.2
Total 100.0% 100.0% 100.0% 100.0%

(Gini coeff=) (.34) (.31) (.50)

—
income. Agricultural income is the value of all crop
production, livestock and beer sales (banana and sorghum),
as well as wages earned in the agricultural sector
(agricultural wage labor). As described earlier, non-farm
income is exclusively that earned from employment in the
non-farm sector.

The Gini coefficients reported in Table 5, indicate
that non-agricultural income is more concentrated in the
hands of high income households (Gini=.50) than is
agricultural income (Gini=.31). Where the wealthiest 15% of
households control 33% of income derived from farm
production and wage labor, they control nearly 50% of non-
farm income. Not only do these findings seem to contradict
our initial expectations that non-farm income would help
equilibrate inequalities in the agricultural sector, but

they suggest that income received from non-farm employment

37
may even compound differentials in agricultural income.
Though we observed earlier in our multivariate analysis of
employment patterns that the number of days worked in the
non-farm sector was inversely related (beta=-.12) to days
worked in agricultural wage labor (an important component of
agricultural income), we did not look at the relationship
between incomes generated from these two sources. Looking
now at this correlation we find that households generating
large agricultural incomes also tend to generate higher
incomes in the non-farm sector (r=.08, significant at .01

level).

Table 6. Percent of Households, Total Income, On-farm
and non-farm by Farm Sise Group

% of % of % of
Farm size % of Total Agric. Non agri.
a o c e o e ncome
< .25 6.7 3.6 2.9 8.6
e25 - e50 19o]. 12.0 1°e8 19e8
e50 - 1.0 32.0 27e5 28e3 21e1
1.0 -' 2.0 2‘e‘ 28o? 30.0 21e1
2.0 - 3.0 11.1 16.4 16.4 16.2
> 3.0 6.7 11.8 11.6 13.2
Total 100.0% 100.0% 100.0% 100.0%
(Gini coeff=) (.19) (.21) (.08)

Since farm size is highly correlated with total
household income (r=.46) one might suppose that the same
pattern of income distribution would hold true for farm—
size categories as for income groups. Table 6 tells us that
while total household income is, indeed, concentrated on the

larger farms (Gini=.19), the distribution of non-farm income

38
is considerably more equitable (Gini=.08) than is
agricultural income (Gini=.21), contrary to the pattern of
findings in the previous table. These findings suggest that
non-farm employment can, in fact, help narrow the income gap
between large and small farmers. But because a small group
of large farmers tend to have higher levels of training,
hold higher paying jobs in the non-farm sector, and treat
their farms as business enterprises, their incomes from this
combination of sources tends to be disproportionately high.
Thus, when comparing income classes (as opposed to a farm
size classes) the effect of non-farm employment is to create

even greater disparities in total income.

CONCLUSIONS

The findings of this study permit us to conclude that
the inheritance and accumulation of landholding can take us
a long way toward understanding income inequalities in
Rwanda, but not all the way. For those whose holdings do
not permit an escape from poverty, their labor endowment may
-- not through agricultural wage labor, but through
employment in the non-farm sector. Households that have
managed to secure both land resources and the skills
necessary to participate in the non-farm sector put
themselves in an exceptional position vis-a-vis the
generation of income and accumulation of wealth. Households
of this type also tend to use their resources to their

fullest advantage; by working off the farm themselves and

39
hiring large amounts of farm labor they maximize income from
farm production.

At the other extreme, we find the near-landless farm
household with little formal education and no training for
alternative employment. Though not the largest identifiable
group, farmers with less than 0.5 hectares of land, no
schooling and no non-farm income are numerous, account for
26% of Rwanda's rural population. Providing for these
households, which, underemployed on their own farms, have
begun to rely on the meager wages they earn as day laborers,
will require heavy investment in the rural non-farm sector,
in particular, in areas where poor and landless households
can be competitive. Yet, since non-farm employment currently
accounts for only 12.7% of total rural income in Rwanda, its
effect on reducing inequality at this stage is minor. As
non-farm employment opportunities begin to grow we can
expect that they may further counterbalance income
inequalities rooted in distribution of landholding.

Specifically, we must begin with a comprehensive study
of rural non-farm employment in Rwanda, such as that
currently under way in the Ministry of Planning. Not only is
it important to survey the individuals and households
engaged in this sector, but also the owners and would-be
owners of the small enterprises that drive this sector.
Those enterprises showing the greatest promise for growth
must be identified. Credit, training and extension programs

supportive of these industries must be reinforced. And

40
equally important Rwanda must look ahead and stimulate
growth in areas that will be responsive to domestic and
international markets of the future.

In the interim, young men and women in Rwanda are
saying that they wish to follow in the path of their parents
and become farmers themselves; yet 85% of these young
people, and the majority of their parents, believe that they
will not inherit enough land for the subsistence needs of
their families (Clay et al., 1989). Among the 85% of young
men in Rwanda who believe they will not inherit enough land
to meet their families' needs, nearly a quarter expect to
have no other options than to make future careers as
agricultural laborers; others expect to migrate in search of
employment; none believes that the strategy of saving their
earnings in order to purchase additional landholding holds
any promise for the future. Those who expect to acquire the
skills necessary to participate in the non-farm sector are
few. Though most parents have an opinion about how their
children should plan for the future, an alarming 26%
indicate that their children will have to make do on their
own.

In the light of these findings, one must conclude that
the primary challenge facing parents, communities and
government officials today will be to overcome the
advantages of land ownership by providing the nation’s less

fortunate young men and women with the skills, access to

41
credit and guidance necessary to build a future beyond the

encroaching hedgerows of their family inheritance.

LI ST OF REFERENCES

LIST OF REFERENCES

Almas, Reidar

1981 "Farm Relief Services-Their Role in Rural
Development." Speielegie_3pzelie, Vol XXXI, no 2.
Anderson, Denis
1982 Small Industry in Developing Countries: A
Discussion of the Issues,” sz1d_neyelpppep§, vol
10, I11

Barlett, F. Peggy
1986 "Part-time Farming: Saving the Farm or Saving the
Life-style?" Repel_§peielegy, Vol 51, Fall, no 3,
Louisiana State University, Baton Rouge, Louisiana

Bilsborrow, Richard E.
1987 "Population Pressures and Agricultural Development
in Developing Countries: A Conceptual Framework

and Recent Evidence.” Weplg Qevelopmepp, 15:2,

Blarel, Benoit

1989 e in c P an u
gentexte de Eezte Preeeipp erographiqpe; L’

Experience dp Rwanda. Banque Mondiale avec la '
Division des Enquetes et Statistiques Agricole du

MINAGRI Rwanda, 1989

Bloom, David E. and Freeman, Richard B.

1986 "The Effect of Rapid Population Growth on Labor
Supply and Employment in Developing Countries."
v ' , #3 Sept.

Byerlee, Derek and Carl K. Eicher
1972 "Rural Employment, Migration and Economic
Development: Theoretical Issues and Empirical

Evidence From Africa." W21

Byerlee, Derek, C.K. Eicher, C. Liedholm, Dunstan S.C.
Spencer
1977 "Rural Employment in Tropical Africa: Summary of
Findings." Africap Renal Economy Progrem, Wopking
Eeper £29, Michigan State University and Njala
University College, Sierra Leone, Department of
Agricultural Economics.

42

43

Byerlee, Derek
1973 ”Indirect Employment and Income Distribution
Effects of Agricultural Development Strategies, a
Simulation Approach Applied to Nigeria." African
Rural Economy Paper #9, Michigan State University.

Byerlee, D. and King R.P.
1977 "Income Distribution, Consumption Patterns and
Consumption Linkages in Rural Sierra Leone."
African Rural Employment Paper # 16, Michigan
State University.

Idem
1978 "Factors Intensities and Locational Linkages of
Rural Consumption Patterns in Sierra Leone."
Jou of cu tur conom cs, vol
60, #2.
Cawley, Mary
1983 ”Part-time Farming in Rural Development, Evidence
from Western Ireland". Sociolegia Ruralis, vol.23,
no 1.

Chuta, Enyina and Carl Liedholm
1985 Employment and Qzeyph in Smell-Seale Industry.

New York: St Martins Press

Clay, Daniel, T. Kampayana and J. Kayitsinga
1989 "Inequality and the Emergence of Non farm
Employment in Rwanda,” Paper presented at the
annual meeting of the Rural Sociological Society,
Seattle, WA.

Clay, Daniel C., J. Kayitsinga, T. Kampayana, I. Ngenzi and
J. Olson
1989 "Strategies Non-Agricole au Rwanda: Rapport
Preliminaire." SESA Document de Travail, Service
des Enquetes et des Statistiques Agricoles,
Rwanda.

Coughenour C. Milton and Wimberley C.Ronald
1989 "Small and Part-time Farmers." In Rumal Society in

The US: Issues for the 19805, edited by Dillman
Don. A. and Hobbs Daryl J., Boulder, Colorado.
West View Press, 1982.

Craig, Jane
1977 n o e a 0 ° e u 'o of bird
World Reform, Oxfam-Canada.

Deb, N.C. and Hossain M.

1984 "Demand for Rural Industries Products in
Bangladesh." The Bemgledesh Qeyelopment Studies,

vol XII, #182

44

D.S.A (Division des Statistiques Agricoles, Ministére de 1’
Agriculture et de l'Elevage)
1987 sc ' ' o a' e e
Qara2teristigues_de_11.8griculture_au_awandao
E.N.A 84 Rapport 2, Fevrier 87.
Durkheim, Emile
1933 The_Dixision_of_Labor_in_Sosietr- Originally
published in 1893. Translated by George Simpson.
New York: The Free Press.

Fisseha, Yacob and Omar Davies
1981 ”The Small-scale Manufacturing Enterprises in
Jamaica: Socioeconomic Characteristics and
Constraints.” MSU Rural Development Series,
Working Paper #16

Fuller, M. Anthony

1983 "Part-time Farming and the Farm Family. A note for
the Future Research". Speiplpgie_gpre11e, vol 23,
no 1.

Goldkind, Victor
1961 "Socio-Cultural Contract in Rural and Urban

Settlement Types in Costa Rica.” Rprel_§peiplegy
vol 26, no 4.

Haggblade Steve, Hazell Peter and James Brown

1987 Non- a. nka es i u a Sub- n
Implieeriepe. Discussion Paper, Report No: ARU 67.
World Bank.

Haggblade, 8., Jacques Defay and Bob Pitman
1979 "Small Manufacturing and Repair Enterprises in
haiti: Survey Results." Michigan State University
Rural Development Series, Working Paper #4

Harper, Malcolm and Soon, Tan Thiam

1979 Small Enterprises in Deyelppipg Qeuptriee [ Case

studies.and.§onclusiona Intermediate Technology
Publications LTD, London.

Hazell, Peter and Aisla Roell
1983 "Rural Growth Linkages: Household Expenditure
Patterns in Malaysia and Nigeria." Inernational
Food Policy Research Institute, Research Report
#41

Hawley, Amos
1950 Hunan_E2o1og2i_A_Theorx_of_99nmunitx_§tru§ture.

New York: Ronald Press.

45

Heffernan, William D., G. Green, R. P. Lasley, and M. F. Nolan
1981 "Part-time Farming and the Rural Community."

Bural_§ooiol_gx vol 46 no 2-

Hekken, van P. M. and H. U. E. Thoden van Velzen
1972 S 't a e n a ia.
Paris: Mouton.

Heusch, Luc
1966 t 'v s ' t a s .
Universite Libre de Bruxelles, Institut de
Sociologie

Hoselitz, B.F.
1956 "Small Industry in Underdeveloped Countries."

Iournal_of_Esonon12_Eiatorx. vol 19

Hymer and Resnick Stephen
1969 ”A Model of Agrarian Economy with nonagricultural
Activities” The American Economic Review #59,
George Banta Co, Wisconsin.

I.L.O (International Labor Office)
1983 "Promotion of Employment and Income for the Rural
Poor, including rural women, through non-farm
activities," Geneva.

I.L.O

1972 Enn1oxnent1_InQone__nd_Equalitxi_a_fitrategx_for
QifiQQV erimg Bred ueriye Emeplpym em; in Kenya. Geneva

I.L.O
1974 De en r mm of Em o e
Egu1tx___d_9____h_f2r_t___2h111221ne§ Geneva

Kagame, Alexis

1972 gm eprege ge l’ethno-piereire dp Rwande, Tome

premier, Butare, Rwanda.

Kempiski, T.
1961 "Rural Migration." Research Notes, Rural

Sociologxi Vol. 26. no 1-

Krasovec, Stane
1983 "Farm Adjustment to Pluriactivity" Seciologia
Rprelie, Vol 23, no 1.

 

Liedholm, Carl and Chuta, Enyina

1976 "The Economics of Rural and Urban Small-Scale
Industries," African Emplpymepr gaper t 1 .

Michigan State University.

46

Idem
1975 ”The Role of Small Scale Industry in Employment
Generation and Rural deverlopment: Initial
Research Results from Sierra Leone." African Rural
Employment Paper #11, Michigan State University

Liedholm, Carl
1973 ”Research on Employment in the Rural non-farm
Sector in Africa.” African Rural Employment Paper
#5 Michigan State University

Liedholm, Carl and Mead, Donald
1987 ”Small Scale Industries in Developing Countries:
Empirical Evidence and Policy Implications"
Michigan State University, Paper # 9, prepared for
the Employment and Enterprise Development
Division, Office of Rural and Institutional
Development, Washington D.C.

Lipton, Michael
1985 Land_As§e1§_and_82ral_£oxertx- The World Bank
Working Papers #744, Washington D.C.

McKenzie Roderik D.
1933 "Industrial Expansion and the interrelations of
Peoples." In Edward B. Reuter (ed.), Rece and

Qplrpre_§pnreere. New York: McGraw Hill.
Maquet, J. Jacques

1961 The_Eren1se_of_1negualitx_1n_8wanda Oxford

University Press, London, Ibadan, Accra

Marx, Karl
1967 Qan11a1_:_A_Qritigue.of.£olitisal.£sononx. Volume
1. New York, NY: International Publishers.

Mead, Donald C.

1989 "Non-farm Income and Food Security: Lessons from
Rwanda" Paper presented at the Michigan State
University Food Security Seminar in Harare,
Zimbabwe.

Mendras, Henri

1970 The_2an1shins_2easant. Cambridge. MA: The MIT

Press .

Narongchai et A1
1983 O - e t hai an
Bangladesh:The Industrial Managemant Co., LTD

47

Nyrop, Richard F., Brenneman Lyle E., Hibs Roy V., James
1985 Charlene A., Macknight Susan and McDonald Gordon
C.

, Fifth ed. United States
Government, Washington D.C.

Okafor, Francis C.
1982 "Environmental Constraints and Part-time Farming
in Southern Nigeria.” geejeprpel, no 4:359-366.

Olson, Jennifer

1990 The_Innast_of_Qhans1ng_Sosioesononi§_£a2tor§_in
' t' e . Department of

Geography, Michigan State University

Rwanda

1967 Ipyeer_ip_3yepge, Ministry of International
Cooperation and Planning, 3rd ed.

Stallmann, Judith I. and James W. Pease
1983 Rural Manufacturing in three Regions of Honduras."
Michigan State University

Steel, Williams
1977 Snall:S2a1e_En21oxnent_and_£rodustion_1n
e 'n 0 'es- ' e o a
Praeger Publisher, New York, N.Y.

World Bank
1990 World Bank Development Report 1990. Washington
D.C. Oxford University Press.