1, + Z, 77
Carl K. Eicher and Carl :L:edholm, Co-editors,
:. .. The MS
S and: the Ne
S on a wide
I non-farm e
; farmers; a
:;. faculty an
;: :... ...::relsearch a
:: The pa
S...and .may re
n ': otific:ati
S.. . ..
* :" .. .. '
' .:. ,. .
... : .. . mm
U Rural Development Paper series is designed to further the
e analysis of rural...development :in Africa, Latiln America, Asia
aar East. The papers will :report research findings onl community:'
t and rural development ini historical perspective :as well .as an
ry rural development programs.. The series.will include paper :
range of topics such .as. alternative rural development strategies;
employmentt and smallTscaTe industry; marketing problems of small
agricultural extension; interrelat:i onships between technology, ..
, and income distributionn; and evaluation of rural development .
While the paperss will convey the. research findi::ngs of MSU
d visiting'schblars, :a few. pape-r will :be published by. re- ..
and policy-makers working with I:MSU scholars on cooperative .
nd action programs in the ..ieid.
pers are aimed at.teachers, researchers, policy-makers, donor
and rural development practitioners. Selected papers will be
into French, Spanish, and.Arab ic. Libraries, individuals,
utions may obtain single. c-ies of the MSU papers free of charge
quest their names be placed on a.mailing list for periodic
ons of published papers by writing to:
MSU Rural Development Papers
Department of Agricultural Economics
206 International Center.
Michigan State University
East Lansing, Michigan 48824
.*t;; .:;.~.:; ;; ..;:. 'lr: ; ;i
RURAL NON-FARM EMPLOYMENT:
A REVIEW OF THE STATE OF THE ART*
*This paper has been developed and published as part of Michigan
State University's Off-Farm Employment Project, which is financed
by the Office of Rural Development and Development Administration,
Development Support Bureau, U. S. Agency for International Devel-
**International Labour Office (Geneva)
***Professor of Economics, Michigan State University
This paper is one of a series of reports produced by Michigan State
University's Off-Farm Employment Project. The project, which is funded
by the Office of Rural Development and Development Administration,
Development Support Bureau, U. S. Agency for International Development,
has the basic purpose of enhancing the ability of AID missions and host
country institutions to identify and implement programs and policies that
generate off-farm employment and income opportunities benefiting the rural
poor. One of the major components of the project is the generation of
new knowledge relating to rural non-farm activities. In collaboration
with host country institutions and AID missions, detailed field surveys
of small-scale enterprises are currently being conducted in Bangladesh,
Jamaica, Honduras, and Thailand; the results of these studies will be
published in this series. A second component of the project involves the
marshalling and dissemination of existing knowledge of rural non-farm
activities. The present State of the Art paper fits into this project
component. This paper, however, should be viewed as only an initial effort
at disseminating and reviewing the existing knowledge, since a major
monograph, building on the State of the Art paper and the results of the
individual country studies, will subsequently appear in this series along
with several other case studies. Previously completed studies in this
area currently available through the Off-Farm Employment Project include:
1. Carl Liedholm, "Research on Employment in the Rural Non-farm
Sector in Africa," African Rural Employment Paper No. 5,
2. Carl Liedholm and Enyinna Chuta,"The Economics of Rural and
Urban Small-Scale Industries in Sierra Leone," African
Rural Employment Paper No. 14, 1974.
3. Enyinna Chuta, "The Economics of the Gara (Tie-Dye) Cloth
Industry in Sierra Leone," February, African Rural
Economy Working Paper No. 25, 1978.
4. Adewale Mabawonku, "An Economic Evaluation of Apprenticeship
Training in Western Nigerian Small-Scale Industry,"
African Rural Employment Paper No. 17, 1979,
5. Steve Haggblade, J. Defay and Bob Pitman, "Small Manufacturing
and Repair Enterprises in Haiti: Survey Results,"
Michigan State University Rural Development Series,
Working Paper No. 4, 1979.
Copies of these papers as well as additional information on the Off-Farm
Employment Project can be obtained by writing:
Off-Farm Employment Project
Department of Agricultural Economics
Michigan State University
East Lansing, Michigan 48824
Several individuals contributed to the production of this monograph
and their assistance deserves recognition. Herb Kriesel actively partic-
ipated at all stages of the project and played a major role in shaping
the final document. Dennis Anderson, Cliff Barton, Steve Haggblade,
Peter Kilby, and Paul Strassmann read earlier drafts and made detailed
and valuable comments that greatly improved the final manuscript. We
also benefited from the comments of the participants attending Michigan
State University's (May 1979) Off-Farm Employment Conference, where a
preliminary draft of this paper was actively discussed. Useful assistance
was also provided by other members of the Off-Farm Employment Project
including: Jim Boomgard, Yacob Fisseha, Annette Francis, Jim Pease
and Judy Stallmann. Special credit for putting all of this in print
under tight schedule goes to Anne Morris, who produced the final manu-
script in record time.
TABLE OF CONTENTS
1. INTRODUCTION . . . . . . . ... . . . 1
2. DESCRIPTIVE PROFILE OF RURAL NON-FARM ACTIVITIES . . . 2
2.1 Importance of Rural Non-farm Activities . . . 2
2.1.1 Primary Employment . . . . . . . . 3
2.1.2 Secondary Employment . . . . . . . 5
2.1.3 Importance of Non-farm Income . . . . . . 6
2.2 Sectoral Composition of Rural Non-farm
Activities . . . . . . . . . . 8
2.3 Equity Implications of Rural Non-farm
Activities . . . . . . . . . . 11
2.3.1 Size of Enterprises Engaged in Rural
Non-farm Activities . . . . . . . . 12
2.3.2 Rural Non-farm Earnings and Wages . . . . .. 12
2.3.3 Relation of Non-farm Activities to
the Landless and Near Landless . . . . .. 14
2.4 Growth of Rural Non-farm Activities . . . ... 16
3. DETERMINANTS OF THE ROLE OF RURAL NON-FARM ACTIVITIES . 22
3.1 Demand for Rural Non-farm Activities . . ... 22
3.1.1 Rural Income . . . . . . . .... . 22
3.1.2 Backward and Forward Production Linkages . . .. 24
3.1.3 Foreign and Urban Demand . . . . . . .. 29
3.2 Supply of Rural Non-farm Activities . . . .. 30
3.2.1 Labor Intensity of Rural Non-farm Activities . 31
3.2.2 Labor Productivity ................ 32
3.2.3 Capital Productivity of Rural Non-farm
Activities . . . . . . . ... ........... 35
3.2.4 Alternative Production Techniques in
Rural Non-farm Activities . . . . . .... .39
3.2.5 Economies of Scale in Rural Non-farm
Activities . . . . . . . . . . 43
3.2.6 Profit, Savings, and Reinvestment Rates
of Rural Non-farm Activities . . . . ... 45
3.2.7 Supply of Entrepreneurship . . . . .... 48
4. MAJOR POLICY AND PROJECT ISSUES . . . . . .... .52
4.1 Introduction . . . . . . . .... . 52
4.2 Policies Resulting in Factor Price
Distortions . . . . . . . . ... 54
4.2.1 Interest Rates . . . . . . . . ... .54
4.2.2 Tariffs . . . . . . . . . . . 56
4.2.3 Foreign Exchange . . . . . . . . 58
TABLE OF CONTENTS (continued)
4. MAJOR POLICY AND PROJECT ISSUES (continued)
4.2.4 Other Tax Incentives . . . . . . .... 58
4.2.5 Minimum Wage Regulations . . . . . .... 59
4.3 Policies with Nonprice Supply Effects ...... 60
4.3.1 Development of Infrastructure . . . . .. 60
4.3.2 Industrial Policies . . . . . . . . 61
4.4 Policies Affecting the Demand for
Rural Non-farm Activities . . . . . 62
4.5 Rural Non-farm Enterprise Project
Issues . . . . . . . .. .. . . 64
4.5.1 Introduction . . . . . . . . .. . 64
4.5.2 General Project Issues . . . . . .... 64
4.5.3 Specific Project Issues . . . . . .... 65
22.214.171.124 Credit Assistance . . . . . . .... 66
126.96.36.199 Technical Assistance . . . . . . .... 72
188.8.131.52 Management Assistance . . . . . .... 75
184.108.40.206 Marketing Assistance . . . . . . .... 76
220.127.116.11 Common Facilities . . . . . . . ... 78
5. SUMMARY . . . . . . . ... ... .. .79
BIBLIOGRAPHY . . . . . . . . ... .. .. .85
LIST OF TABLES
PERCENTAGE OF RURAL LABOR FORCE WITH PRIMARY
EMPLOYMENT IN RURAL NON-FARM ACTIVITIES . .
SHARE OF NON-FARM INCOME IN TOTAL RURAL
HOUSEHOLD INCOME . . . . . . .
SECTORAL COMPOSITION OF RURAL NON-FARM
EMPLOYMENT IN SELECTED COUNTRIES . . .
SIZE OF LAND HOLDING AND RELATIVE IMPORTANCE
OF NON-FARM INCOME IN TOTAL HOUSEHOLD INCOME
GROWTH OF RURAL NON-FARM EMPLOYMENT:
SELECTED COUNTRIES . . . . . . .
SIZE OF ENTERPRISE AND LABOR INTENSITY (K/L)
IN SELECTED COUNTRIES . . . . . .
SIZE OF INDUSTRIAL ENTERPRISE AND CAPITAL
PRODUCTIVITY (Q/K) IN SELECTED COUNTRIES .
CAPITAL PRODUCTIVITY AND LABOR INTENSITY BY
PROCESS AND LOCATION FOR SELECTED ACTIVITIES
A COMPARISON OF OFFICIAL AND NONOFFICIAL RATES
OF INTEREST IN SOME DEVELOPING COUNTRIES . .
AVERAGE TARIFF RATES BY END-USE GROUPS
FOR SOME DEVELOPING COUNTRIES . . . . .
TYPES OF ASSISTANCE, THEIR VARIOUS FORMS
AND DELIVERY CHANNELS . . . . . . .
. . . 4
. . . 7
. . . 9
. . . 18
. . . 33
. . . 55
. . . 57
. . . 67
International donor agencies and the governments of many developing
countries have recently begun to devote increasing attention to the
development of policies and programs for expanding productive employment
and earnings opportunities in the various rural non-farm activities under-
taken in developing countries. This growing interest stems from and
reflects the increased international concern for equity and employment
objectives, and the corresponding reduction of emphasis on the earlier
strategies that had focused primarily on growth and output objectives.
The de-emphasis of growth and output objectives reflects a disillusionment
with the inequitable results of rapid growth in certain countries and
the disappointing results of the attempts to rapidly industrialize by
establishing large-scale, urban-based, capital-intensive industries. In
a number of developing countries, not only was the overall rate of growth
quite low, but employment in the industrial sector failed to keep pace
with population growth and, in some cases, even declined in absolute terms.
Unfortunately, there have been few empirical or analytical studies of
rural non-farm economic activities in developing countries. The excellent
World Bank paper on rural enterprise and non-farm employment (World Bank,
1978a) notes, for example, that "there is little concrete evidence" on
many of the important characteristics of these activities, and Morawetz
(1974, p. 525), in his recent review of the literature, states that
"remarkably little is known about its composition and characteristics."
As a result, those charged with formulating and executing rural non-farm
programs and policies are generally forced, of necessity, to make decisions
"unencumbered by information."
The present paper is an attempt to fill the information lacuna
relating to rural non-farm economic activities in developing countries.
Although the paper, of necessity, concentrates on rural non-farm
activities, one must continually keep in mind that these activities
represent only one facet of the rural development process. The
first section provides a descriptive profile of rural non-farm
activities and sets forth the most important issues relating to their
nature, extent, and composition. The second section examines the deter-
minants of their role in development and focuses on factors influencing
the demand for and supply of these activities. A final section examines
the major policy and program issues relating to rural non-farm activities.
2. DESCRIPTIVE PROFILE OF RURAL NON-FARM ACTIVITIES
There are several important sets of issues relating to the extent and
nature of rural non-farm activities. These are mainly descriptive issues,
which can usefully serve to provide a foundation for understanding the
role of rural non-farm activities in developing economies. The issue areas
are: 1) the quantitative significance of rural non-farm activities;
2) their sectoral composition; 3) equity implications; and 4) growth
2.1 Importance of Rural Non-farm Activities
One of the first issues to be considered is whether or not non-farm
activities are quantitatively an important component of the rural economy.
Given the paucity of comprehensive income and value added statistics
relating to rural areas of most developing countries, one must, of necessity,
rely primarily on employment data for illumination of this issue. The
importance of non-farm activities as a source of primary employment in
rural areas will be examined first, followed by a consideration of the
importance of these activities in providing secondary or part-time employ-
ment; the relative importance of non-farm income is then discussed.
2.1.1 Primary Employment
The evidence available from national censuses and various regional and
rural surveys indicates that non-farm activities provide an important
source of primary employment in the rural areas of most developing countries.
In the vast majority of the eighteen developing countries where relatively
recent data on the subject are available, one-fifth or more of the rural
labor force is primarily engaged in non-farm activities (table 2.1).
Although the rural non-farm percentage ranged from 14 to 49 percent, in
over three-quarters of the countries the percentage fell between 19 and 28
The figures provide a minimal estimate of the magnitude of primary
employment in rural areas. First, they generally reflect the employment
charactersitics of the rural villages with populations below 5,000; if the
larger rural towns were included, the rural non-farm percentage would
likely be larger.I Second, there are certain measurement errors that
operate to cause systematic undercounting of non-farm activities. In some
African countries rural respondents will claim farming to be their main
See, for example, the evidence cited in the World Bank, 1978a. The
dividing line between "rural" and "urban" is arbitrary, particularly in
census data collected in most countries. They are often framed in terms
of urbanization characteristics rather than minimum size or occupational
structure size and, consequently, settlements of a few thousand are often
classified as "urban." The U.N. definition of "urban" is localities with
20,000 or more inhabitants. This broader definition, which includes small
and medium sized towns, is used in this paper.
RURAL LABOR FORCE WITH PRIMARY EMPLOYMENT
IN RURAL NON-FARM ACTIVITIES
Percentage of Rural
Labor Force Primarily
Country Year Coverage Employed in Non-Farm
Guatemala 1964 All rural 14%
Thailand 1970 All rural 18
Sierra Leone 1976 Male-rural 19
South Korea 1970 All rural 19
Pakistan 1970 Punjab only 19
Nigeria 1966 Male-3 dist. W. State 19
India 1966 All rural 20
Uganda 1967 Four rural villages 20
Afghanistan 1971 Male-Paktia Region 22
Mexico 1970 All-Sinaloa State 23
Colombia 1970 All rural 23
Indonesia 1971 All rural 24
Venezuela 1969 All rural 27
Kenya 1970 All rural 28
Philippines 1971 All rural 28
W. Malaysia 1970 All rural 32
Iran 1972 All rural 33
Taiwan 1966 All rural 49
1. Guatemala: World Bank (1978a)
2. Thailand: Thailand (1973)
3. Sierra Leone: Byerlee, et al.(1977)
4. Korea: Korea (1972)
5. Pakistan: World Bank (1978a)
6. Nigeria: Mueller and Zevering (1970)
7. India: World Bank (1978a)
8. Uganda: Brandt, et al. (1972)
9. Afghanistan: Gerken (1973)
Mexico: World Bank (1978a)
Colombia: World Bank (1978a)
Indonesia: Leiserson (1974)
Venezuela: World Bank (1978a)
Kenya: I.L.O. (1972)
Philippines: I.L.O. (1974)
West Malaysia: World Bank (1978a)
Iran: Dhamija (1976)
Taiwan: Ho (1976)
occupation even if they engage only part-time in this activity. In
addition, women's participation in non-farm activities is often not
counted as employed labor even when these activities result in trans-
2.1.2 Secondary Employment
These primary employment statistics also understate the magnitude
of rural non-farm activities, because they fail to reflect those farmers
who engage in non-farm activities on a part-time or seasonal basis.
Data on secondary employment are not generally available for most countries.
The limited evidence indicates that from 10 to 20 percent of the rural
male labor force undertake non-farm work as a secondary occupation. In
Western Nigeria, for example, 20 percent of the rural males engaged in
non-farm work on a part-time basis, while in Sierra Leone, Afghanistan,
and Korea, the figures were 11, 16, and 20 percent, respectively.1
There are significant monthly variations in the amounts of rural
farm and non-farm employment over the agricultural cycle. Farm and non-
farm employment move in opposite directions. There is no period when
non-farm employment disappears and, thus, non-farm employment does compete
with farm employment during periods of the peak agricultural demand. Data
from Nigeria reveal that the peak in non-farm labor use is nine times that
in the slack periods (Norman, 1973). The fluidity of labor between a
number of activities on a seasonal basis is thus a striking feature of
For sources, see table 2.1.
In summary, non-farm activity in rural areas thus appears to provide
a source of employment for from 30 to 50 percent of the rural labor force
in the developing nations, when primary and secondary occupations are
included. Consequently, in terms of employment, non-farm activities are
quantitatively an important component of the rural economy2 that should
not be overlooked in the design of rural development policies or programs.
2.1.3 Importance of Non-farm Income
In view of the magnitude of rural non-farm employment, it is not
surprising that non-farm activities also provide an important source of
income for rural households. Although data on rural incomes are generally
lacking for most countries, the evidence from those countries where
information is available indicates that non-farm earnings account for
over one-fifth of total rural household income (see table 2.2). Indeed in
Sierra Leone, where a detailed rural household survey was recently under-
taken, non-farm income was found to provide 36 percent of rural household
income, while in Taiwan the comparable figure was 43 percent.3
1There is evidence that the figure may be as high as 50 percent in
some countries. Luning (1967), in a survey of rural villages in Northern
Nigeria, reports that 48 percent of the employed males engaged either
full or part-time in rural non-farm activities, while Norman (1973) reports
that, in the same area, 47 percent of male labor time is devoted to these
Several recent historical studies have revealed that non-farm
activities were important, amounting from one-third to two-thirds of
rural occupations, in rural areas of Europe during the 16th and 17th
centuries. See, for example, MacFarland (1977, p. 157) for U.K. and
DeVries (1974) for Holland.
3Moreover, in the Philippines (I.L.O., 1974, p. 504), 37 percent of
the rural households derived their main sources of income from non-farm
SHARE OF NON-FARM INCOME IN TOTAL RURAL HOUSEHOLD INCOME
Country Year Percentage
Korea 1975 22%
Pakistan (5 villages) 1968 23
Northern Nigeria (3 villages) 1974 28
Sierra Leone 1974 36
Taiwan 1975 43
Japan 1975 70
Korea: Korea (1975)
Pakistan: Kuhnen (1968)
Northern Nigeria: Matlon (1977)
Sierra Leone: Unpublished results of African Rural Employment
Taiwan: Taiwan (1976)
Japan: Japan (1976)
2.2 Sectoral Composition of Rural Non-Farm Activities
Another set of issues revolves around the sectoral composition of
rural non-farm activities. Specifically, what are the types of activities
undertaken within the sector, and which of these are quantitatively the
There is a wide array of activities being undertaken within the rural
non-farm sector. In terms of the Standard Industrial Classification
categories, the most important components are manufacturing, services,
and commerce activities. This composition is revealed in table 2.3, in
which data are presented on the breakdown of primary employment in rural
areas for selected developing countries. Manufacturing ranges from 22 to
46 percent, commerce ranges from 11 to 35 percent, while services range
from 10 to 50 percent of total rural non-farm employment.1 Other non-farm
activities, such as construction, transport, and utilities, generally
account for less than 25 percent of rural non-farm employment.2
The relative importance of rural, as opposed to urban, manufacturing
may appear somewhat surprising. There is empirical evidence to indicate
that employment in small, rural manufacturing enterprises often exceeds
that in large urban manufacturing firms. In Sierra Leone, 86 percent of
the total manufacturing sector employment and 95 percent of the manufacturing
establishments were located in rural areas (Liedholm and Chuta, 1976).
The percentage of rural manufacturing employment in other countries ranges
1Agricultural processing and marketing activities would be reflected
in these figures; fishing and livestock activities would not.
2The exception is Afghanistan, where transport (particularly camel
driving) is an important rural non-farm activity.
SECTORAL COMPOSITION OF RURAL NON-FARM EMPLOYMENT IN SELECTED COUNTRIES
Afghanistan India Indonesia Leone ippines Korea Colombia Malaysia Taiwan
(1970) (1966) (1971) (1975) (1970) (1970) (1970) (1970) (1966)
Manufacturing 46% 39% 29% 40% 34% 30% 33% 22% 27%
Construction 9 14 5 2 11 10 8 5 4
Trade and Commerce 11 14 34 35 15 24 19 22 13
Services 10 24 27 23 30 29 33 41 50
Other 24 9 5 -- 10 7 7 10 6
100% 100% 100% 100% 100% 100% 100% 100% 100%
NOTES: Includes utilities, transport,
and miscellaneous; omits "other and unknown."
Afghanistan: Gerken (1973)
India: World Bank (1978a)
Indonesia: Leiserson (1974)
Sierra Leone: C. Liedholm and E. Chuta (1976)
Philippines: I.L.O. (1974)
Korea: Korea (1972)
Colombia: World Bank (1978a)
Malaysia: World Bank (1978a)
Taiwan: Ho (1976)
from 70 percent in Bangladesh (Bangladesh Institute of Development Studies,
1979), 63 percent in Malaysia (World Bank, 1978b), 57 percent in India (World
Bank, 1978b), to 32 percent in Korea (Korea, 1972). These figures may actu-
ally understate the true magnitude of rural manufacturing activity because
country censuses often fail to pick up the very small rural enterprises.
The Sierra Leone small industry survey found that rural manufacturing
employment had been underestimated in Sierra Leone by almost one-half
(Liedholm and Chuta, 1976). A recent pilot rural industry survey in
Bangladesh indicated that, in one rural district, the number of rural
firms was twenty times greater than indicated by the official statistics
(Ahmed, Chuta, Rahman, 1978).
Within the rural manufacturing component of the rural non-farm sector,
there is a surprising diversity of activities undertaken. The most im-
portant activity in the majority of countries appears to be clothing
production followed by wood working, metal working, and food processing.
Clothing production, for example, accounted for 53 percent of the rural
manufacturing employment in Sierra Leone, 41 percent in Korea, 24 percent
in Taiwan, 32 percent in Western Nigeria, and 52 percent in rural Bangladesh.
Several alternative, sometimes conflicting, classification schemes
have been developed that reflect and highlight the differing forms or
subsectors of activities within the rural non-farm sector. The "informal
sector," a concept popularized by the I.L.O. (1972), refers to activities
that "operate largely outside the system of government benefits and
regulations" and are characterized particularly by ease of entry, small
scale of operation, family ownership, and unregulated and competitive
markets; the "formal sector" activities are essentially the obverse of
these "informal activities."1 Within the manufacturing sector, a distinction
is often made between "artisan activities," where production is completely
under the direction of the owner, and "small factories," where there is
a greater division of labor and the manager, rather than the artisan, is
the central figure (Staley and Morse, 1965, p. 7). A common dividing line
between the two is frequently ten workers. Finally, the World Bank (1978b)
distinguishes "artisan and informal enterprise" from "modern small
enterprise having perhaps ten to fifty workers."1 The available evidence
indicates that the vast majority of the existing rural non-farm enterprises
in developing countries would fall in the "artisan and informal enterprise"
2.3 Equity Implications of Rural Non-farm Activities
The next set of issues focus attention on the equity implications of
rural non-farm activities. Specifically, what is the size of rural non-
farm enterprises, what is the relative income earned by those engaged
in these activities, and finally, what is the relation of rural non-farm
activities to the landless and near landless?
1Steel (1978) recommends a tripartite division between: (1) "casual
and home production," where the marginal product of labor is zero, there
are only family employees, and there are no barriers to entry; (2) the
"intermediate sector," where there are barriers to entry making the marginal
product of labor positive; and (3) the "modern" sector, where there are
severe barriers to entry.
2See below, p.12.
2.3.1 Size of Enterprises Engaged in Rural Non-farm Activities
The size of the enterprises engaging in rural non-farm activities is
an issue of some interest. Variations in size are particularly important
for ascertaining the equity implications of rural non-farm policies or
The available empirical evidence is limited but does indicate that
the vast majority of rural non-farm activities are undertaken by very
small-scale, artisan and informal enterprises. In Sierra Leone, the
average rural industrial firm employed 1.6 workers, and 99 percent of the
firms employed less than 5 individuals (Liedholm and Chuta, 1976). In
rural Jamaica (Davies, Fisseha, Francis, and Kirton, 1979), the average
rural enterprise engaged 1.8 workers. In rural Western Nigeria, an I.L.O.
survey reported that the average industrial firm engaged 2.6 workers (Mueller,
et al., 1970). The results from a similar survey in rural Bangladesh revealed
the average rural enterprise employed 3.8 workers, inclusive of proprietors,
and 84 percent of the enterprises engaged fewer than 6 workers (Bangladesh
Institute of Development Studies, 1979). These findings indicate that
most rural enterprises are of a very small size and thus may be potentially
an important target group for policy makers concerned with the poor.
2.3.2 Rural Non-farm Earnings and Wages
A related issue centers on whether or not the earnings from rural
non-farm occupations or the average incomes of non-farm rural households
"Small-scale" is not a precisely defined concept. There are at
least 50 different definitions used in 75 countries. (See, for example,
Staley and Morse, 1965, and Georgia Tech., 1975). As a working definition
for this paper, "small-scale" is defined to include those establishments
employing less than 50 persons.
are above those in agriculture. This issue is of particular importance
given the increased concern with income distribution, and more specifically,
the widespread interest in identifying the sectoral characteristics of
the rural and urban poor.1
The rather limited amount of available data suggest that, on the
average, the wages and incomes generated by rural non-farm activities
generally exceed those generated by farming. In Sierra Leone, for example,
the income per consumer equivalent for rural non-farm households was $155
while that for farming households was $125.2 In Malaysia, the differential
was even larger, with the income per family member for non-farm households
amounting to $402 and that for land-abundant padi farmers amounting to
$230 (Bell and Hazell, 1976). The lack of detailed data for other areas,
however, makes it difficult to assert with complete confidence that such
differentials necessarily exist in all countries. Moreover, these figures
are averages and mask important variations by season, region, sex, education,
skill level, and type of employer (see, for example, Byerlee, et al., 1976).
There is preliminary evidence to indicate that there is a positive
association between non-farm income sources and income level in some rural
areas. In Northern Nigeria, for example, non-farm income comprised 20
1See for example, Chenery (1974, p. 19).
Data from Africa Rural Employment Study Sierra Leone. The data
include the imputed value of home production that is consumed.
Similar results have been reported from studies in Kenya, Tanzania,
Mexico, Tunisia, and Northeast Brazil. In Kenya, the average earnings
per annum for adult wage workers on small farms was $106 while that for
rural nonagricultural enterprises was $126 (I.L.O., 1972, p. 77). The
average household income of rural nonagricultural households was 53% higher
in Mexico, 41% higher in Tanzania, and 22% higher in Tunisia than the
average incomes of farm households in these countries (Van Ginnekan, 1976,
p. 41). For Brazil, see World Bank (1978) p. 82.
percent of the total household income of the lowest income decile, but
rose to comprise 37 percent of the income of the highest decile (Matlon,
1977, p. 80). Correspondingly, iii Sierra Leone, non-farm income was
28 percent of income in the lowest decile and 37 percent in the highest.1
Although total non-farm income appears to be somewhat concentrated in
the higher income groups in the rural areas, average rural non-farm
earnings are still substantially below earnings in the urban areas. In
both Kenya and Sierra Leone the average rural non-farm earnings are sub-
stantially below the statuatory minimum wage.2 Those engaged in rural
non-farm activities thus are for the most part an important component of
2.3.3 Relation of Non-farm Activities to
the Landless and Near Landless
Non-farm activities are particularly important for those rural house-
holds with little or no land. Indeed in countries for which data exist,
there is a clear negative relationship between the importance of non-farm
activity and farm size. As farms become smaller, the share of non-farm
income in total household income becomes larger (see table 2.4). In
Sierra Leone, for example, rural households cultivating more than 15 acres
earned less than 20 percent of their income from non-farm sources, while
those cultivating less than 1 acre earned more than 64 percent from such
Sierra Leone, Rural Employment Study, preliminary results.
In Kenya, the annual earnings from nonagricultural rural wages was $126
while a male subject to the minimum wage would earn $297. In Sierra Leone,
even the small-scale proprietor earned an annual return 13 percent below
that earned by an employee in a large-scale urban enterprise (Liedholm and
SIZE OF LAND HOLDING AND RELATIVE IMPORTANCE OF
NON-FARM INCOME IN TOTAL HOUSEHOLD INCOME
Size Non-farm Income Share Total
of in Total House- Household
Country Holding hold Income Income
Korea 0.00 1.23 acres 42% $ 495
171 1.24 2.47 acres 19 724
2.48 3.70 acres 10 1015
3.71 4.94 acres 10 1309
4.95+ acres 8 1781
Taiwan (Taichung Region) 0.00 1.19 acres 59 652
197 1.20 2.39 acres 44 764
2.40 3.58 acres 33 1136
3.59 4.40 acres 27 1200
4.41 7.18 acres 40 1811
7.19+ acres 15 1989
Sierra Leone 0.00 1.00 acres 64 587
1.01 5.00 acres 51 404
174 5.01 10.00acres 32 546
10.01 15.00acres 26 770
15.00+ acres 17 927
Northern Nigeria 0.00 2.46 acres 57 479
197 2.47 4.93 acres 31 377
4.94 7.40 acres 26 569
7.41 9.87 acres 15 769
9.88+ acres 24 868
1. Korea: World Bank (1978a)
2. Taiwan: Taiwan (1970)
3. Sierra Leone: Unpublished data from African Rural Employment
4. Northern Nigeria: Matlon (1977)
sources. Clearly non-farm earnings provide a significant portion of
the total income of those rural households with little or no land.
Moreover, these non-farm earnings can be sufficiently large in some
instances to enable landless or near landless rural households to generate
a total household income greater than that of the larger-sized farms.
In Sierra Leone the total income of those rural households with less
than an acre of land was $587, while the income of those households with
1 to 5 acres was $404. A similar pattern was found to exist in Northern
Nigeria.1 Such results call into question the notion that farm size is
consistently an accurate measure of total household income or is con-
sistently a good indicator of who are the rural poor. In much of Africa,
where land is not a primary limiting factor, there does not appear to
be a positive relationship between land holding and total income in
the smallest land holding categories. Such a relationship, however,
may hold in much of Asia, where land does tend to serve as a primary
constraint to income generation.2
2.4 Growth of Rural Non-farm Activities
The final descriptive issue is whether or not rural non-farm activities
and employment decline in importance as development proceeds. More spec-
ifically, do rural non-farm activities decrease as rural incomes rise and
opportunities for trade increase? On this issue, there has been some
divergence of views.
This relationship is also found in an even more extreme form in
Japan. Disposable income per capital in 1974, in index form, is 111 for
farms of 0.1 0.5 ha., 99 for 0.5 1 ha., 89 for 1 1.5 ha., 88 for
1.5 2.0 ha., and 94 for 2.0 ha. See Kato and Izumida (1977, p. 3).
See Korea and Taiwan, for example, in table 2.4.
The issue was sparked by the 1969 paper, "A Model of an Agrarian
Economy with Non-Agricultural Activities," (1969) by Stephen Hymer and
Stephen Resnick. In this paper, they develop a model of the rural economy
in which rural non-farm activities, denoted as Z goods, are hypothesized
to decline as rural incomes rise and opportunities for trade increase.
Resnick, in a subsequent article (1970), provided empirical evidence for
the contention by tracing the decline of rural industry in Burma, Philippines,
and Thailand from 1870 to 1938. Comprehensive time series data were not
available, however, and Resnick, of necessity, was forced to rely on frag-
ments of evidence from various sources. Consequently, the results of
the study, while interesting, cannot be considered conclusive.
The empirical evidence available for more recent periods would in-
dicate that rural non-farm activity and employment recently have been
increasing, rather than decreasing, with development. Table 2.5 presents
the figures from ten countries for which aggregate time series data exist.
In all cases, the rates of rural non-farm employment growth were positive,
ranging from 3.2 percent per year in Korea to 9.4 percent per year in
Taiwan.1 These results, while not conclusive, would indicate that rural
non-farm activities and employment have been increasing in absolute terms
over time in developing countries.
There is evidence to indicate that, in most areas, non-farm employ-
ment has been growing more rapidly than farm employment. Dennis Anderson,
using secondary I.L.O. data, has shown that the rural labor force increased
Two qualifications should be noted. The growth of non-farm employ-
ment in rural towns which are rapidly growing centers of such employment -
is omitted. On the other hand, part of the "recorded" increase in rural
non-farm employment may be monetization of activities previously undertaken
in the household.
GROWTH OF RURAL NON-FARM EMPLOYMENT: SELECTED COUNTRIES
Rural Non-farm Non-farm Share of
Employment Growth Rural Labor Force (%)
Country Period Rate (per year) Initial End
(%) Period Period
Taiwan 1955-66 9.4% 30% 49%
Kenya 1969-74 8.8 NA NA
Philippines 1961-71 8.5 NA NA
Mexico 1960-70 5.6 14 23
Indonesia 1961-71 5.5 17 24
Iran 1956-72 4.8 20 33
India 1953-60 4.0 NA NA
Afghanistan 1964-71 3.9 NA NA
Korea 1960-74 3.2 18 19
Taiwan: Ho (1976)
Kenya: Steele (1976)
Philippines: Gibb (1974)
Mexico: Mexico (1978a)
Indonesia: Leiserson (1974)
Iran: Dhamija (1976)
India: India (1965)
Afghanistan: Gerken (1973)
Korea: Korea (1976)
faster between 1959 and 19701 than the agricultural labor force in all
regions except Latin America. In addition, the specific country data
presented in table 2.5 reveal that over time the percentage of the rural
labor force engaged in non-farm work has risen. Finally there is some
cross-sectional evidence that there is a positive association between
the share of the rural labor force engaged in non-farm work and the level
of per capital income. These results indicate that rural non-farm
activities, rather than decreasing, are becoming a more important source
of employment in rural areas.
Although the available evidence indicates that aggregate employment
and output in rural non-farm activities have been increasing, an important,
related issue is: What has been the growth performance of the individual
types of rural non-farm activities? There are many heterogeneous kinds
of activities covered by the rural non-farm umbrella and some of these
activities might be expected to have declined while others might be
expected to have increased over time.
Unfortunately, much of the available information is antecdotal or
episodic. Comprehensive time series data for particular types of rural
non-farm activities are generally not available, although some information
on specific manufacturing activities do exist for some countries such as
the Philippines (Gibb, 1974; Anderson, 1979), Sierra Leone (Liedholm and
Chuta, 1976), and Haiti (Haggblade, et al., 1979). Among the major types
World Bank (1978a). In Latin America from 1960-70, the agricultural
labor force increased 0.8 percent per year while the rural labor force in-
creased 0.6 percent per year. (These results are based on preliminary
figures from the I.L.O. and must be interpreted with some caution.)
The non-farm data in table 2.1 were regressed against levels of per
capital income yielding the following result: Percent of rural labor force
in non-farm activities = -.12 + 4.06 log per capital income. R = .2, F = 4.
of existing activities the available evidence indicates that tailoring,
dress making, furniture making, baking, and rice milling have continued
to grow in importance even after large-scale, domestic factory production
of these commodities has begun. Shoe production, leather production, and
pottery appear to have generally declined in importance.1 A mixed record
appears with blacksmithing, and spinning and weaving. It should be noted
that the kinds of activities undertaken by some of the important artisan
groups have been evolving. In some countries, for example, rural black-
smiths, who previously were primarily engaged in the production or servicing
of hand tools, now also produce or service animal-drawn or mechanized farm
equipment, and irrigation equipment (Liedholm and Chuta, 1976; Child and
Kaneda, 1975). Moreover, several newer types of artisan activities, such
as bicycle, auto, and electrical repair activities have grown particularly
rapidly in recent years. These newer activities reflect the increased
service-oriented nature of many artisan activities as the level of income
and urban factory production increase. In addition, certain types of craft-
oriented artisan activities designed for the international market, such as
gara (tie-dye) cloth in Sierra Leone (Liedholm and Chuta, 1976) and wood
carving in Haiti (Haggblade, et al., 1979) have also been growing rapidly
in certain countries. Finally, a few "modern" factory activities, some of
which have emerged from smaller enterprises, such as metal working factories
in India (Berna, 1960) and cement block production and essential oils
Additional evidence on the decline of these particular activities is
found for India (Prasad, 1963), Ethiopia (Karsten, 1972), and Burma (Resnick,
2Spinning and weaving has declined in the Philippines and Sierra Leone,
but has increased, since Independence, in India.
(luxury perfume) production in rural Haiti (Haggblade, et al., 1979),
have also begun to increase in importance.1
These differential growth patterns are important to recognize in
the design of programs and policies for the rural non-farm area. Govern-
mental policies, particularly with respect to large, modern industries
and agriculture, influence growth patterns of individual activities
within each country.2 Although some of the existing rural non-farm
activities will evolve and new activities will emerge, the sheer magnitude
of these existing informal artisan activities in most countries indicates
that any major transformation will take many years to complete. Stewart
(1977) has estimated that it will take several decades before the "formal"
sector will begin to absorb even the additions to the labor force in most
developing countries. Consequently, attention must continue to be directed
towards enhancing many of the types of activities represented in the
existing structure of rural non-farm enterprise, even if, in the longer
run, many of them will eventually decline in importance or disappear.
For an excellent listing of the types of "modern" small enterprises,
both urban and rural, likely to increase in importance, see Staley and
Morse (1965, p. 97ff). Locational, process and market influences are
In India hand loom production declined from 1901-1948 under
colonial rule (Prasad, 1963; table 14), but increased after Independence
with government encouragement.
3Investments in most "informal" rural non-farm enterprises, for
example, would be fully amortized within a ten to twenty year time
3. DETERMINANTS OF THE ROLE OF RURAL NON-FARM ACTIVITIES
Both the future and existing patterns of rural non-farm activities
are determined by a set of factors influencing the demand for and supply
of these economic activities. These demand and supply issues will be
examined in the next section.
3.1 Demand for Rural Non-farm Activities
There are three principal sources of demand for the products and
services of rural non-farm activities that should be considered. The
primary source is the demand generated from the incomes of rural consumers.
A second source of demand arises from the backward and forward production
linkages with the agricultural and large-scale industrial sectors. The
final, important source of demand is provided by the urban and foreign or
export sector. The issues relating to each of these demand sources will
now be examined.
3.1.1 Rural Income
A central issue is whether or not the demand for rural non-farm
activities should be expected to increase as rural incomes increase.
There have been some divergent views expressed on this particular matter.
Hymer and Resnick (1969) have argued that rural non-farm goods and
services, "Z goods," are "inferior" goods and thus the demand for these
goods will decline as rural incomes rise. Pack, in his review (1977) of
the report of the 1972 I.L.O. mission to Kenya, also contends that
increasing the incomes of the poor may have, at best, only a limited
effect on the demand for the goods and services of the "informal sector,"
which includes rural non-farm activities. Neither Hymer and Resnick, nor
Pack, present any empirical evidence to support this view and Pack himself
admits, "relatively little is known about the magnitudes involved."
Mellor (1976), Liedholm and Chuta (1976) and various I.L.O. Employ-
ment Missions (1972), (1974), have contended that there is a strong,
positive relationship between rural income and the demand for rural non-
farm activities. The available evidence, though limited, tends to
support this view. Virtually all the standard analyses of rural
household expenditure surveys undertaken in such diverse countries
as India (Mellor, 1976), Kenya (Massell, 1969), and Uganda (Massell and
Parnes, 1969), indicate that the income elasticity of demand by rural
households for nonfood consumption items is positive and, in most cases
exceeds unity, and that these activities account for an increasing pro-
portion of a rural household's budget as its income rises.
Although these analyses are indicative, they are not conclusive,
because they fail to differentiate between those nonfood consumption items
produced in the rural areas and those produced in urban areas or imported.
King and Byerlee's (1978) pioneering rural expenditure survey in Sierra
Leone, however, does differentiate activities by origin or location and
reveals that the rural expenditure elasticity for rurally-produced non-farm
consumption activities is 1.4 (i.e., indicating that an increase of rural
incomes of ten percent raises the expenditure for rural non-farm goods
and services by fourteen percent). Studies by Leurquin in Ruanda-Urundi
(1960, p. 313) and Gibb in the Gapan area of the Philippines (1974) also
indirectly indicate that an increasing share of income is allocated to
rural non-farm activities as income rises. Consequently, these few studies
reveal that rural non-farm goods are not "inferior" (i.e., possess an
expenditure elasticity below zero), and rather than being viewed as an
over-riding constraint, the demand induced from increasing incomes should
be viewed as a strong force for the growth of rural non-farm activities
in developing countries.1 Clearly, further research is needed to verify
these relationships in other countries and the magnitude of the elas-
ticities for individual rural non-farm activities.
3.1.2 Backward and Forward Production Linkages
A second major demand issue centers on the nature and extent of the
production linkages between rural non-farm activities and other sectors
of the economy, particularly the agricultural and large-scale industrial
sectors. Specifically, there are the "forward" linkages from the rural non-
farm sector, where rural non-farm outputs serve as inputs to other sectors,
and the "backward" linkages from the rural non-farm sector, where this
sector provides a demand for the output of other sectors. There are diverg-
ing opinions and varying empirical evidence on the production linkage issue.
In this section, the rural non-farm sector's linkages with the agricultural
sector will be examined first, followed by a discussion of its linkages
with large-scale industry.
A. 0. Hirschman, in his classic book, The Strategy of Economic Develop-
ment (1958), contends, without detailed empirical evidence, that the linkages
between agriculture and other sectors are quite weak. Yet, Mellor (1976),
argues that linkages with agriculture are, or could be, potentially quite
significant; indeed these agricultural linkages are an essential ingredient
in Mellor's "rural-led strategy of development."
The composition and magnitude of these effects depends importantly on
the pattern of agricultural growth. See below (p. 62) as well as Mellor
(1976), and Johnston and Kilby (1975) for further discussions of these effects.
The empirical evidence on rural non-farm linkages with agriculture
tends to be somewhat limited. The vast majority of the input-output studies
fail to include any explicit rural non-farm activities, and thus they mask
or understate the rural non-farm linkages with agriculture.1 The few
input-output studies that specifically include rural non-farm activities,
however, indicate the "forward" and "backward" production linkages from
this sector to agriculture are often quite important. Such results are
found in input-output studies of India (Krishna, 1973; Falcon, 1967); of
the Muda River Area of Malaysia (Bell and Hazell, 1976); of the Philippines
(I.L.O., 1974) and of Sierra Leone (Byerlee, et al., 1977). Additional
support for the strength of these "production linkages" is also found in
several detailed industrial case studies undertaken in these and other
With respect to the "forward linkages" from rural non-farm activities
to agriculture, the empirical studies indicate that rurally produced
agricultural inputs are particularly important where traditional "inter-
mediate" agricultural technologies are utilized. Johnston and Kilby's
(1975) analysis of farm equipment in India, Pakistan, and Taiwan stresses
that traditional tools are most often made by rural artisans, while improved
implements, and irrigation pumps and motors are likely to be fabricated
by light engineering workshops located in rural towns.2
IThe rural non-farm activities are either omitted, often for lack of
data, or are lumped together with agriculture or "modern" large-scale
Tractors, combines, and other large items with high performance
characteristics, as well as fertilizers, tend to be produced abroad or
in urban areas. Consequently, the nature of the agricultural technology
adopted has important effects on rural non-farm linkages (see below,
p. 64 for a further discussion).
Karsten's study of rural blacksmiths in Ethiopia (1972) and Liedholm
and Chuta's analysis of rural artisans in Sierra Leone (1976) provide
further support for the crucial role played by rural artisans in providing
inputs for traditional agriculture in Africa. Liedholm and Chuta note
that approximately one dollar of rural blacksmithing output, particularly
in the form of hoes, knives, and axes, is demanded for every one hundred
dollars of agricultural output. Moreover, both Child and Kaneda's (1975)
analysis of diesel tubewell production in Pakistan and Cartillier's (1975)
study of electric tubewell manufacturing in India point out the extensive
growth of these light engineering activities in those rural areas where
improved agricultural practices have been adopted.
With respect to the "backward" linkages from rural non-farm activities
to agriculture, the weight of the empirical evidence suggests that these
are quite significant. Most of the studies focus on the linkages between
rural agricultural processing and the agricultural sector, although rural
transport and rural marketing activities are also potentially important
backward linkages. Krishna's (1973) input-output study of India indicates
that such activities as the processing of gur, tobacco, sugar, cashew nuts,
and flour have among the highest intersectoral linkages. Unfortunately,
the location and size of these activities is not specified. Falcon's (1967)
study of agricultural-industrial inter-relationships in Pakistan, however,
reveals that crop flows to small-scale processing activities, the majority
of which are rural, are more than five times the flow to urban, large-
Indeed the strength of this "backward" linkage from rural non-farm
processing to agricultural production depends crucially on the choice and
location of the processing technology involved. Although there is some
indication that a range or mix of technologies will sometimes be optimal,
most of the case studies of processing indicate that small-scale, rurally-
based processing activities generally are economically efficient in
developing countries.1 Studies of rice processing in Indonesia (Timmer,
1975) and Sierra Leone (Spencer, 1976) reveal the significant links between
small, rural rice mills or hand-pounding and rice production. Similar
results for palm oil processing in Nigeria are reported by Miller (1965).
In summary, these various empirical studies indicate the importance of
backward and forward linkages of the rural non-farm sector with agriculture
and point to the need for future researchers to incorporate explicitly rural
non-farm activities when analyzing sectoral interactions.
The empirical evidence on the production linkages between rural non-
farm activities and large-scale industry is also somewhat sparse. Only
a few of the input-output studies, as noted previously, explicitly include
rural non-farm activities, and case studies in this area are very limited.
The "forward" production linkages from rural non-farm enterprises
to large-scale industry, where these enterprises provide intermediate or
capital goods to large-scale industry, are most frequently discussed in
terms of subcontracting relationships (see World Bank 1978b). Such sub-
contracting arrangements have been particularly important in Japan, where
approximately 60 percent of all small-scale units are subcontractors to
large firms (Vepa, 1971; Paine, 1971). There is little evidence on the
1See below (p. 35) for a more detailed examination of production
location of these subcontracting enterprises, however, and it is difficult
to specify how much of these subcontracting activities are undertaken by
small, rural enterprises. Subcontracting to small enterprises does not
appear to be as extensive outside of Japan, although the empirical evidence
is admittedly very sketchy. In India, for example, subcontracting repre-
sents about 1 percent of the total product of small-scale industry (Vepa,
1971). In Africa and Latin America, subcontracting and the "forward"
linkages from small, rural non-farm enterprises to large enterprises are
minimal (Liedholm and Chuta, 1976; Pack, 1978; UNIDO, 1969).2
Indeed the "backward" production linkage from small-scale, rural,
non-farm activities to large-scale industry appears more extensive than
the "forward" linkage. Even in Japan output flows from large to small
enterprises are almost three times as much as that from small to large
enterprises (Hoselitz, 1968). Several examples of small, rural non-farm
enterprises purchasing their intermediate inputs from larger firms are
cited by Pack (1977) for Kenya, Child and Kaneda (1975) for Pakistan,
and Liedholm and Chuta (1978) for Sierra Leone.3 In summary, the nature
and extent of these production linkages from rural, non-farm activities
There is also subcontracting undertaken for foreign firms for com-
ponent production or assembly. These relationships are found extensively
in Hong Kong, Taiwan, Mexico, Singapore, and a few other areas; most of
their subcontractors are urban based (see Watanabe, 1976). Some foreign
initiated rural subcontracting does exist, however, in a few developing
countries such as Haiti, but these relationships have not been studied.
2See below, p. 64 for a discussion of the policy issues relating to
the promotion of subcontracting activities.
3These backward linkages may have fewer indirect employment effects
than the forward linkages (see Pack, 1977).
have not been extensively examined, and additional, detailed studies
of these intersectoral linkages are clearly required.
3.1.3 Foreign and Urban Demand
The final important demand issue centers on the nature and magnitude
of the foreign and urban demand for the products of rural non-farm
enterprises.1 The lack of detailed data on the location of productive
activities in most countries makes it difficult to derive any definitive
conclusion on this issue.
The available evidence does indicate that rural non-farm products
do enter into international markets and that, for some activities, the
international market is a major component of the total market. In Iran,
handicrafts, including carpets, is the largest export item after oil
and 60 percent of the handicraft activity is undertaken in rural areas
(Dhamija, 1976). In India, handicraft and handloom commodities account
for approximately 6 percent of the country's value of exports (Government
of India, 1965). Finally, evidence from Sierra Leone indicates that
approximately one-fifth of the total production of the rural gara (tie-dye)
industry is exported (Liedholm and Chuta, 1976).
The evidence also supports the view that the international market
is or is potentially an important component of demand for rural non-farm
activities. The only detailed study in this area has been undertaken by
Huddle and Ho (1972), who examined the international demand for eighty-
one different "culturally-oriented" products. Specifically, their study
reveals that the overall income elasticity of demand exceeds one in high
The actual and potential demand of government for rural small enter-
prise goods and services (e.g., school uniforms) should also be considered.
income countries for products such as wood carvings, brassware, and
earthenware. Thus the overall demand for these products should be
expected to increase importantly as income in high-income countries
increases. More detailed studies focusing on the location of these
production activities would be useful in determining whether or not
there are differences in the elasticities of demand between rurally-
produced, "culturally-oriented" commodities. Little is known of the
factors determining the proportion of the total market captured by each
individual country. Such studies are important for designing effective
programs and policies enabling developing countries to take full ad-
vantage of this important market.1
3.2 Supply of Rural Non-farm Activities
Supply factors also play crucial roles in determining both the
current and future nature, extent, and composition of rural non-farm
activities. The important supply issues relate primarily to the efficiency
with which rural non-farm enterprises utilize their economic resources,
both in static and dynamic terms. The key static efficiency issues
center on the existing labor intensity, capital productivity, factor
substitutability, and economies of scale in rural non-farm activities;
the dynamic issues relate primarily to the potential for the expansion
of capital and entrepreneurship in these activities.
Policies such as those designed to improve the rural infrastructure
and the marketing system may be crucial for fully exploiting this inter-
national market. See below, p. 60 for further discussion of these policy
3.2.1 Labor Intensity of Rural Non-farm Activities
One important supply issue is whether or not rural non-farm activities
are more labor-intensive than other segments of the economy. Since in
most developing countries, capital and foreign exchange are relatively
scarce, and labor, particularly unskilled, is relatively abundant, those
activities and techniques of production that are more labor-intensive would
generate the largest amount of employment per unit of scarce factor and
thus appear to represent activities or technologies most "appropriate" to
their factor endowments.
The evidence available indicates that existing rural non-farm activities
are generally more labor-intensive than other segments of the economy. Most
of the studies utilize the capital-labor ratio or its reciprocal to measure
labor intensity and this labor intensity measure must be interpreted with
caution. Moreover, the majority of these studies have either examined
specific processes within industries or compared large and small enterprises
and thus generally have not specifically compared rural non-farm activities
with other activities. Since the great bulk of rural non-farm enterprises
are very small and utilize certain types of processes, it is possible to
impute cautiously from the results of these more general studies.
There have been studies involving at least nine countries comparing
the labor-capital ratio, or its reciprocal, of large- and small-scale
enterprises. In every country, the smaller scale enterprise group
possess a higher labor-capital ratio (or lower capital-per-worker ratio)
1See White (1978), and Morawetz (1974) for a general discussion.
Such factors as excess capacity, heterogeneity of capital and labor,
stock versus flow problems, and use of market prices have been mentioned
(see Bhalla, 1975).
than the larger scale enterprise group, with the labor intensity of
small firms ranging from four to fifteen times higher than the large
firms (see table 3.1).
Although these studies reveal that smaller enterprises are more
labor-intensive than larger enterprises, they do not differentiate
between rural and urban enterprises and thus do not conclusively verify
whether rural non-farm activities are themselves more labor-intensive.
There is one exception that sheds some light on the issue. The Sierra
Leone small enterprise survey (Liedholm and Chuta, 1976) does differ-
entiate by location and indicates that the small-scale, rural enterprises
in Sierra Leone are at least twice as labor-intensive.1 Although
additional studies in this area are needed, the analysis points to the
relative labor intensity of rural non-farm activities.
3.2.2 Labor Productivity
A second supply issue centers on how the labor productivity of
rural non-farm activities compares with labor productivity in other
segments of the economy. The relevance of this issue, however, is
perhaps as important as the conclusions derived from the various
During the 1950s and 1960s several international study groups and
productivity missions equated "efficiency" with labor productivity;
The differences were statistically significant.
2See, for example, the I.L.O. report cited in Kilby (1962).
SIZE OF ENTERPRISE AND LABOR INTENSITY (K/L) IN SELECTED COUNTRIES
Country' Size of Enterprise 1-10 workers 11-50 workers 50+ workers
Fixed Capital Per Worker ($)
Japan (1966) $ 934 $1040 b $4333
India (1965) 278 557 2450
("small") ("medium") ("large")
Malaysia (1968) 521 997 2671
Philippines (1970) 1020 2850 8000
("small") ("medium") ("large")
Sierra Leone (1974) 158 225 1175
Kenya (1960) 772 986 3108
Ghana (1970) 1372 3724 6468
Mexico (1970) 3700 9500 14,500
("small") ("medium") ("large")
Colombia 3000 --- 13,400
Japan: Okhawa and Tajima (1976)
India: World Bank (1978b)
Malaysia: Okhawa and Tajima (1976)
Philippines: World Bank (1978b)
Sierra Leone: Liedholm and Chuta (19i
Kenya: I.L.O. (1972)
Ghana: Steel (1977)
Mexico: World Bank (1978b)
Colombia: World Bank (1978b)
aFixed capital per worker ($)
bNumbers in brackets refer to
size distribution when
they differ from heading
consequently, enterprises with high average labor productivities were
considered the most "efficient."1 The efficacy of this view has rightly
been questioned by many observers (see White, 1978). Although the
level of labor and managerial skills do affect the labor productivity
measure, the amount of capital with which each employee works is a
crucially important determinant of labor productivity. A low labor
productivity figure for a rural non-farm establishment may be a reflec-
tion, as White notes, "of the efficient combination of labor with low
levels of capital in developing countries" (White, 1978, p. 30). The
average productivity of labor thus would not appear to be a very useful
efficiency criterion, particularly if labor is not viewed as the binding,
scarce resource constraint.2
The available empirical evidence generally indicates that the average
productivity of labor is lower in small-scale enterprises than in the
larger scale enterprises (see the results of the studies listed in table
3.1). The results from Sierra Leone (Byerlee, et al., 1979) indicate
that small, rural, non-farm enterprises have somewhat lower labor pro-
ductivities than do their larger scale counterparts in urban areas.
Such findings are not surprising in light of the results presented in the
previous section that the larger enterprises possess greater amounts
of capital per worker.
1In a dynamic framework, the World Bank (1978a) argues that, "gains
in labor productivity are essential." The crucial issue is, however, how
the increase in output (the numerator) is to be maximized given the scarce
If any labor measure is to be used, it would be the marginal rather
than the average product of labor. Specifically, the value of the marginal
product would be equated to the wage in any efficiency measure. Skilled
labor or managerial labor might be a scarce resource in certain industries.
3.2.3 Capital Productivity of Rural Non-farm Activities
A third supply issue revolves around whether or not rural non-farm
enterprises use the scarce factor, capital, as efficiently as do other
enterprises or activities. There have been divergent opinions expressed
on this issue.
Several international groups and individuals, including Nicholas
Kaldor, argued during the 1960s that small-scale, labor-intensive
activities would use not only more labor, but also more of the scarce
factor, capital, than their larger scale counterparts. Hence they
argued that these small-scale, labor-intensive activities would possess
lower output-capital ratios and would be consequently less efficient
than the larger, more capital-intensive enterprises. There could be a
conflict between the objectives of maximizing output and employment if
such a condition exists. (See Baer and Herve, 1966; Morawetz, 1974). During
the 1970s several sources (World Bank, 1978a; World Bank, 1978b; Pack, 1974;
Liedholm and Chuta, 1976; Marsden, 1969) have argued that small-scale
labor-intensive enterprises might also be more efficient in the utilization
of capital (i.e., possess higher output-capital ratios) than their more
capital-intensive counterparts. There need not be a trade-off in such a
case between output and employment objectives.
The available empirical evidence relating to the capital productivity
of rural, non-farm activities, while indicative, is not conclusive. Most
of the studies compare either large and small enterprises or processes
within activities and thus do not explicitly focus on rural, non-farm
1See, for example, Robinson (1965), Barber (1969), I.L.O. Report in
Kilby (1962) and I.L.O. (1961).
activities. There are also measurement problems, particularly for the
smallest enterprises, as well as several conceptual problems relating
to the comparison of large and small enterprises. Thus any aggregate
comparison must be interpreted with extreme caution.
The majority of the empirical evidence does appear to provide some
support for Marsden's (1969) contention that "the smaller enterprises,
with a lower level of investment per worker, tend to achieve a higher
productivity of capital than do larger, more capital-intensive enter-
prises." An examination of table 3.2, where the relationship between
the output-capital ratios and size of establishment in the nine countries
where sufficiently detailed results are available, reveals that in only
one country, India, does the overall capital productivity of the largest
enterprises appear to exceed that in the small enterprise. India may
not even be an exception since several other studies (see World Bank,
1978b and Shetty, 1963) found that the capital productivity of larger
enterprises in India is lower than that found in smaller Indian firms.2
1Data from the smallest enterprises (i.e., one through nine worker
category) are very unreliable since the majority do not keep books and
memory recall is limited (see below, p. 50). Aggregate data comparing
large and small enterprises mask the differences in capital productivity
between individual lines of activity. The appropriate, but less readily
available, comparison is between firms producing the same product with
the same degree of vertical integration (White, 1978). See below, p. 42
for size comparison by roughly comparable product groups. Finally,
the most desirable comparison between small and large enterprises would
involve the marginal rather than the average product of capital, but
unfortunately, such data are generally not available; thus average
products must serve as rough proxies. A related issue in a dynamic
or investment framework is how rapidly the marginal productivity of
capital declines with additional capital invested in rural small enter-
See also the debate and exchange on this issue in India between
Mehta (1969) and Sandesara (1966, 1969).
SIZE OF INDUSTRIAL ENTERPRISE AND CAPITAL PRODUCTIVITY (Q/K)a
IN SELECTED COUNTRIES
Country Size of Enterprise 1-10 workers 11-50 workers 50+ workers
Value Added Per Unit of Fixed Capital
Japan (1966) 1.55 3.32 b 1.50
India (1953) 0.10 0.47 0.73
(1-19) (20-49) (500-999)
Pakistan (1960) 1.16 0.37 0.28
Malaysia (1968) 2.01 1.32 1.02
Philippines (1960) 0.96 0.98 1.11
(5-19) (20-49) (500+)
Sierra Leone (1974) 3.20 1.50 0.72
Ghana (1976) 0.60 0.30 0.60
Kenya (1972) 5.60 2.60 1.10
(0-4) (20-29) (100+)
Mexico (1965) 1.34 0.64 0.61
(1-5) (16-25) (500+)
Japan: Okhawa and Tajima (1976)
India: Okhawa and Tajima (1976)
Pakistan: Ranis (1961)
Malaysia: Okhawa and Tajima (1976)
Philippines: Okhawa and Tajima (1976)
Sierra Leone: Liedholm and Chuta (1976
Ghana: Steel (1977)
Mexico: Okhawa and Tajima (1976)
value added per unit of
bNumbers in brackets refer to
size distribution when they
differ from heading
CRefers to distribution
1) activities only
Since the majority of rural non-farm enterprises fall into the
smallest size category (i.e., one through nine) the relative capital
productivity of this size enterprise group is of particular interest.
Several sources (World Bank, 1978b; Okhawa and Tajima, 1976) have con-
tended that the capital productivity of the "very smallest enterprise group"
(i.e., one through nine) might be lower than the next largest small-
scale size category (i.e., ten to fifty). An examination of table 3.2
reveals that such a result appears to hold only for Japan and perhaps
India; thus the capital productivity of the smallest size enterprise,
where most of the rural non-farm enterprises are found, generally compares
favorably with that of the larger-sized enterprises.
Direct evidence on the relative capital productivity of rural non-
farm activities is available only in the Sierra Leone industry study
(Liedholm and Chuta, 1976). This survey indicates that the output-capital
ratios for rural non-farm enterprises are markedly higher than those of
their urban, small-scale counterparts. These various results, while
certainly not conclusive, thus do tend to indicate that rural non-farm
enterprises generally are not only more labor-intensive, but also may
generate more output per unit of capital than their larger scale counter-
parts; thus there may not be any output-employment trade-off, at least
in a static sense. More detailed case studies of rural non-farm enter-
prises are required before more conclusive judgements can be made.
These results by themselves do not provide a sufficient condition for
additional investment in rural non-farm enterprises. Analytically, data on
the marginal productivity of capital as well as information on how rapidly
the marginal productivity would decline with additional capital are required.
Other potentially scarce resources, such as skilled labor, management, and
foreign exchange, should ideally be included in a measure of total scarce
factor productivity rather than simply capital productivity; these inputs
should be examined at their social prices (see, Morawetz, 1974).
3.2.4 Alternative Production Techniques
in Rural Non-farm Activities
Another important issue is whether or not there is an array of
efficient alternative processes in use or available for the types of
non-farm activities undertaken in rural areas. This issue has significant
policy relevance because it provides an indication of whether changes or
distortions in factor prices have any effect on the optimal production
technique or factor proportions. An examination of the efficiency of
alternative processes within individual industries, rather than the
broadly defined aggregates such as large versus small-scale establishments,
is central to resolving the output-employment trade-off issue.
A central theme among engineers and even some economists, particularly
during the 1950s and 1960s, was that few, if any, efficient alternatives
to the capital-intensive processes of the developed countries existed.1
This contention has been strongly attacked in recent years by a wide
array of individuals and organizations.2
The accumulated, empirical evidence supports the view that a wide,
although not unlimited, number of alternative processes exist within most
lines of manufacturing activities. One of the empirical approaches has
been to measure, by means of econometric estimation procedures, the
elasticity of substitution between capital and labor in a number of
activities. These studies have been carried out in at least 25 developing
See sources mentioned in footnote one on the preceding page, as well
as Francis Stewart (1972) and Eckhaus (1955).
2See White (1978) and Morawetz (1972) for a listing of these views.
See ibid. for a good summary of the evidence.
countries and the vast majority find that estimates of the elasticity
of substitution are positive and tend to bunch between values of 0.5
and 1.2. Although these studies indicate that efficient factor substi-
tution is possible, they all suffer from several methological difficulties.
Consequently, these results must be interpreted with some degree of caution.
A second empirical approach has been to delineate, through individual
case studies, the main production techniques and processes in a given line
of activity. The factor-intensity and efficiency of both existing and
potential techniques are then examined. There have been detailed studies
of only a few products or processes. The results from these studies con-
sistently indicate that factor substitution is possible and the difference
in factor ratios can be quite substantial (White, 1978, p. 34). Unfor-
tunately, many of the case studies do not directly include rural non-farm
processes or products.
The evidence from the limited number of case studies explicitly in-
volving rural non-farm products or processes does indicate that there is
factor or process choice within many lines of rural non-farm activity.
In the various case studies of Sierra Leone rural and urban manufacturing
there were at least five clothing processes, six bread processes, five
gara (tie-dye) processes, three metal-working processes, and five rice
milling processes delineated, all of which possessed different factor
1Morawetz (1974), Gaude (1975), and O'Herlihy (1972).
proportions. Similar results have been reported elsewhere by Bhalla
(1965) and Timmer (1972).2
The few case studies involving rural non-farm activities have re-
vealed that rural non-farm processes are generally both more labor-
intensive and more productive per unit of capital than their larger,
often urban-based counterparts in the same industry. Studies in
Indonesia (Timmer, 1972) and Sierra Leone for rice processing (see table
3.3), indicate that both traditional handpounding and small, rural rice
mills were more labor-intensive and generated more output per unit of
capital than the larger scale mills. For cloth production, several
studies of spinning and weaving in India (Sen, 1968; Bhalla, 1964, Raj,
1957) indicate that the output-capital ratios for traditional hand loom
spinning and weaving activities, the majority of which are rural, are
higher than those for factory production. Similar results are found
for clothing production in Sierra Leone (see table 3.3). Finally, for
bread production, rural bakers using traditional, mud ovens are found to
be more labor-intensive and more productive with capital than the larger,
urban-based bakers (see table 3.3). Similar results are found indirectly
See Chuta (1978), Liedholm and Chuta (1976), and Spencer, et al., (1976).
Although there is evidence that several differing techniques and
processes are being used within many lines of rural non-farm activity, the
range of choice is undoubtedly far from complete. There is thus scope for
the development of new processes or products, particularly those that in-
crease overall factor productivity or will be, at least, more labor-intensive.
For an excellent, recent review of the evidence on technical progress, and
research and development in developing countries, see White, 1978. See also
Marsden (1971), and Strassmann (1968).
In India Bhalla (1965) found that handpounding possessed a lower
output-capital ratio (capital productivity) than machine-milled rice.
Small mills, however, possessed higher output-capital ratios than did
the large mills.
CAPITAL PRODUCTIVITY AND LABOR INTENSITY BY PROCESS AND
LOCATION FOR SELECTED ACTIVITIES
Labor-Capital Ratio Output-
Activity/Country (man-hours per $ Capital
of capital) Ratio
A. Rice Milling
1. Sierra Leone
Rural handpounding 638.00 40.90
Rural small-steel-roller mill 1.25 1.80
Urban large rice mill 0.12 1.20
Rural handpounding NA NA
Small rice mill 0.14a 2.60
Large rice mill 0.02a 0.80
Rural handpounding 1.70a 1.20
Small rice mill 0.04a 2.20
Large rice mill 0.04a 1.90
B. Spinning and Weaving
1. India Spinning
Rural, traditional cotton spinning 4.70a 0.15
Factory spinning 0.02a 0.11
2. India Weaving
Rural fly-shuttle hand-loom NA 4.50
Automated power loom NA 0.30
C. Clothing Sierra Leone
Rural tailor, small-scale non- 16.60 8.30
electric sewing machine
Urban tailor, small-scale electric 4.30 2.60
Urban clothing factory, large-scale 2.20 1.70
D. Bread-making Sierra Leone
Rural baker, small-scale, 38.00 19.00
Urban baker, small-scale, 5.30 3.20
Urban baker, large-scale, 2.60 2.60
Rice milling: Sierra Leone: Spencer, et al. (1976); Indonesia: Timmer (1972);
India: Bhalla (1965).
Spinning and weaving: India-spinning: Bhalla (1964); India-weaving: Sen (1968).
Clothing: Byerlee, et al. (1979).
Baking: Chuta (1979)
NOTE: aNumber of workers per $100 of capital stock
for sugar processing in India (Baron, 1975). These studies, while
certainly not conclusive, do indicate in several lines of activity at
least there are rural small-scale processes that generate more employ-
ment and output per unit of capital than their larger scale, urban-
based counterparts; consequently, in these cases, employment-output
conflict would appear to vanish.1
3.2.5 Economies of Scale in Rural Non-farm Activities
A related issue is the extent of the economies of scale in the
existing or potential lines of activity engaged in by rural non-farm
enterprises. This issue is important in determining whether scale
economies are so predominant that policies to emphasize small, rural
activities might result in a loss in economic efficiency or in a rapid
elimination of rural enterprises as markets expand.
The empirical evidence on this issue, though limited, indicates
that the importance of the scale factor varies importantly by type of
industry or activity. There are certain lines of activity where scale
effects appear to be important such as in chemicals, petroleum refining,
and brewing, where surface areas and volume relations become dominant
features, or metal finishing where the setting-up costs of a production
run become important (White, 1978; Scherer, 1970). There are only a
limited number of empirical studies that have verified the existence of
significant scale economies in developing countries. For metal
These findings must be interpreted with some caution since they
are subject to many of the same measurement and conceptual problems
described in the previous subsection. In addition, Kilby (1964) and
Stewart (1977) have argued that even within the same industry there are
important differences in product quality and that these quality differences
are related to the size and capital intensity of the enterprise. It is not
clear, however, that these quality differences are sufficiently large to
vitiate the results.
machinery (Boon, 1976) in Mexico, and cement block manufacture (Stewart,
1976) in Kenya, recent empirical studies have shown that "there are
appreciable economies of scale and that capital-intensive methods are
necessary to capture these economies."
Several other studies have indicated a number of lines of activity
where scale economies do not appear to be significant. In Sierra Leone
results of production function analyses revealed that there was no
evidence of economies of scale in cloth making, wood working, metal
working, and baking (Liedholm and Chuta, 1976). Similar results were
found in India for textiles (Murti and Sastry, 1957) and in Pakistan
for textiles, light engineering, plastics, and leather (Ranis, 1962).
A similar finding was reported for machine goods production in Pakistan
(Child and Kaneda, 1972). Finally, Pack (1974), using world-wide
cross-section data, found no evidence of significant scale economies in
the bicycle, wheat milling, tire, and woolen yarn textile industries. Thus
the limited available evidence indicates that economies of scale do not
appear to be significant in a wide range of activities where rural non-
farm enterprises are involved.
The results of these various empirical studies indicate that many
small rural non-farm enterprises are relatively efficient in their
utilization of existing resources.2 It is also important to ascertain
whether these activities are efficient in a dynamic sense.
See also Pack and Todaro (1969) for a general discussion of scale
economies in the capital goods industry in developing countries.
These findings do not suggest that these rural non-farm activities
are permanently superior to large, urban-based activities or that they
merit support to the exclusion of urban-based, larger scale activities.
Certain types of activity may be better served by large-scale enterprises.
Moreover, both small and large enterprises can and do co-exist and often
serve different markets.
3.2.6 Profit, Savings, and Reinvestment Rates of
Rural Non-farm Activities
An important dynamic efficiency issue centers on the relative profit,
savings, and reinvestment rates of rural non-farm activities. This
capital growth issue is significant for policy because it sheds light
on the economy's growth rate and whether or not there might be output-
employment conflicts over time.1 Galenson and Leibenstein (1955) and
Sen (1968), argue that even though efficient, small-scale, labor-intensive
activities or processes might exist, large capital-intensive activities
or processes should be chosen because they generate the most savings
and reinvestment and, consequently, generate the most rapid growth of
output and employment over time. This outcome occurs, they argue,
because such activities or processes produce the highest returns (profits)
to capital and capital owners have higher savings and reinvestment rates
than do workers. If these contentions are true, then rural non-farm
activities, which tend to be small and labor-intensive, would appear to
provide less growth support for the economy than others.
There are three empirical components bearing on this capital growth
issue that must be examined. The first relates to the relative savings
propensities from profit and employment income; the second relates to
the relative rate of profit per unit of capital generated by rural non-
farm activities as opposed to those generated by others; the last relates
to the relative savings and reinvestment propensities of these activities.
1Power (1962) used the following formulation to relate the components
of growth: (Y/K)(S/Y) = S/K. This criterion was used to judge the con-
tribution of small and large units.
For some of these components, the empirical evidence is available and
indicative, but for others it is limited and not conclusive.
The empirical evidence on the savings propensities from employment
and profit sources is perhaps the strongest. Studies by Houthackker (1961),
Williamson (1968), as well as others reviewed by Mikesell and Zinser (1973),
do tend to support Galenson and Leibenstein's contention that savings out
of labor income is low while that out of profits tends to be high.
The evidence relating to the returns to capital or profit rates
is much more limited, but is also somewhat indicative. Accurate data on
profit rates of small, rural non-farm enterprises are very difficult to
collect and, consequently, there are only a few studies that can shed
light in this area. Nevertheless these few studies indicate that the
profit rates of small, rural non-farm enterprises are substantial. Child
(1973) found that the median gross profit rate for his sample of rural
industries in rural Kenya was approximately 75 percent. Liedholm and
Chuta (1976) report that the average "economic profit" of rural small-
scale enterprises in Sierra Leone ranged from 20 percent for tailoring
to about 200 percent for gara (tie-dyeing). Finally, Huddle (1977)
reported that the financial profit rates generated by a sample of rural
artisans in Colombia exceeded 150 percent. In all these instances, the
rates of return to capital were markedly higher than rates of return
generated by their larger scale, capital-intensive, urban-based counterparts.2
1The opportunity cost of the proprietor's and family's labor has been
There is evidence from urban-based surveys, such as those undertaken
for Karachi (Ranis, 1961) and Delhi (Dhar, 1961), that small-scale, labor-
intensive enterprises generate a higher rate of return to capital than their
large-scale, capital-intensive counterparts.
Thus Galenson and Leibenstein's contention that the profit rates of
capital-intensive enterprises would exceed those of the smaller, labor-
intensive enterprises does not appear to be supported by the limited
empirical evidence to date. Clearly more studies of the profitability
of small, rural enterprises are needed.
Finally, the evidence relating to the savings and reinvestment
propensities of various enterprises is the most scanty and least reliable
of the empirical components. The only empirical studies of the savings
propensities of small enterprises have been of urban enterprises (Ranis,
1961; Dhar, 1958), and the quality of their savings and reinvestment data
for the smallest enterprises was admittedly very poor. Both authors
concluded that "the percentage of total profits which the small scales
are capable or willing to save (i.e., one-half to two-thirds) is not
significantly smaller than that in other scales" (Ranis, 1961, p. 20).
For the small, rural-based enterprises, the majority of which are family
owned activities, the profits from the enterprise often become mixed with
savings and expenditures from other household activities; thus it becomes
very difficult to isolate the savings and reinvestment rates for those
small rural enterprises that are merely a component of a more complex
rural household. Huddle, in a recent study of Colombian artisans (1977),
however, did find that the average savings propensity of the smallest,
self-employed artisan household was 16 percent, a rate double that of
the general population. There is evidence that a large part of the
initial capital and reinvested capital from small, rural enterprises is
derived from savings. For rural non-farm enterprises in Sierra Leone
(Liedholm and Chuta, 1976) 60 percent of the initial capital came from
agriculture or business, while over 90 percent of the expansion capital
came from reinvested profits. These few limited studies indicate that
rural non-farm enterprises may have a savings and reinvestment potential
that is not markedly below that of their larger counterparts. Clearly
more detailed and careful surveys are needed in this area. In summary,
the empirical evidence does not provide support for Galenson and Leibenstein's
contention that the profit, savings, and reinvestment rates of the kinds
of activities that are undertaken by rural non-farm enterprises will
necessarily result in less output and employment for the economy over time.2
3.2.7 Supply of Entrepreneurship
Another dynamic efficiency issue centers on how responsive the supply
of entrepreneurship3 in rural non-farm activities is to changing conditions
over time. This issue is of importance for policy because it provides an
indication of the nature and extent of any constraints to an expansion
of entrepreneurial supply and the efficacy of various policies and programs
designed to overcome them.
Similar results on the importance of self-financing are reported for
small-scale urban enterprises in India (Singh, 1963; Berna, 1960) and for
small and medium size urban enterprises in Nigeria (Harris, 1970).
Data on the rate of growth of individual rural enterprises are
generally lacking and thus the total relationship between profits, savings,
reinvestment, and firm growth cannot be easily examined. Results from
Haiti (Haggblade, et al., 1979) indicate that the average net rate of equip-
ment investment by individual small-scale enterprises was 6 percent per year
between 1974 and 1978. Moreover, the previously described rapid growth of
rural small-scale enterprise employment in general is also somewhat indica-
tive; even if the growth occurs primarily via the establishment of new,
small-scale enterprises, it could be argued that the savings is being used
to establish new, small enterprises rather than to expand existing ones.
Although there are numerous definitions of entrepreneurship, a common
theme is that the entrepreneur is a key decision maker. As such, the entre-
preneur can be treated as a factor of production where performance is
determined by supply factors (Harris, 1970).
There are contrasting views concerning the responsiveness of the
supply of entrepreneurship over time. Economists such as Harris (1970),
Papanek (1971), and Leibenstein (1969) argue that the supply of entre-
preneurship is responsive and that any deficiencies are either transient
or due to market or policy imperfections. Indeed a recent review article
by Leff (1979) argues that it is a "slack variable" and not a crucial
constraint to development. Kilby (1971), as well as those stressing
the importance of either psychological (e.g., Hagen, 1962; McClelland,
1961) or sociological (e.g., Weber, 1930; Cochran, 1965) theories of
entrepreneurial supply are much more pessimistic about the entrepreneurial
supply responsiveness. The relative importance of the determinants of
entrepreneurial supply and success, such as ethnicity, status, education
(both formal and non-formal), are central to the validation of these
Unfortunately, there have been very few detailed studies of rural
entrepreneurs that focus on the determinants of entrepreneurial supply
in developing countries (see Broehl, 1978 for India, and Liedholm and
Chuta, 1976 for Sierra Leone). Most of the studies examine enterpre-
neurship in somewhat larger, urban-based firms and thus their findings
are only of limited usefulness.2
A common finding of all of these studies is that there are generally
serious deficiencies in the entrepreneur's managerial and technical
He does not include the management function in his analysis, however,
and focuses on the innovation and risk-taking functions.
2See, for example, those in India of Berna (1960), Singh (1963),
Gadgil (1959), and Lamb (1955); Kilby (1971) and Harris (1970) in Nigeria;
Marris and Somerset (1973) in Kenya; Carroll (1965) in the Philippines; and
Cochran (1959) in Puerto Rico.
performance in most countries (see Kilby, 1971 for a listing and review).
A crucial question then is whether or not the provision of training
would enable this supply constraint to be overcome. Kilby (1971) and
Cochran (1959) argue, however, that these marginal shortcomings are
"enduring impediments rooted in sociological variables" that cannot be
overcome by training, while Harris (1967) and others argue that appropriate
training would be sufficient to expand entrepreneurship.
The empirical evidence on this issue is somewhat mixed. With respect
to formal education, for example, Kilby (1965), Harris (1970), and Liedholm
and Chuta (1976) have found little or no evidence that formal education
and entrepreneurial success are related. One explanation for this finding
is that nonformal education may be a more relevant form of education than
the formal kind for smaller rural enterprises. In most rural areas the
apprenticeship system is the primary vehicle for providing technical
training; indeed in Sierra Leone 90 percent of the proprietors had
previously served as apprentices. Other nonformal methods of training
are also available in rural areas. One aspect of this training that
appears crucial is record keeping, a skill that can be imparted by non-
formal methods. Yet, most rural entrepreneurs do not keep any records.
The Sierra Leone study (Liedholm and Chuta, 1976), indicates that those
rural enterprises that maintain even a rudimentary set of books are more
successful (i.e., generate more profits) than those that do not. This
same study also reveals that another nonformal type of education, number
of years of on-the-job "experience", is also a significant determinant
of a rural enterprise's success.
In Sierra Leone, only 17% of the proprietors kept any books; in
rural Bangladesh only 6% kept books (Bangladesh Institute of Development
Studies, 1979), while in Jamaica only 11% of the rural proprietors kept
any records (Davis, Fisseha, Francis, and Kirton, 1979).
The nature of the managerial and technical constraint on the
entrepreneurial supply, however, may vary by the type of expansion
envisaged. The supply of entrepreneurial services can be enlarged
either through an expansion of existing enterprises or through a pro-
liferation of new enterprises. There is some evidence that there may
be some deterioration in managerial performance as very small enter-
prises expand (Harris, 1970; Kilby, 1969; Marris and Sommerset, 1972).
Harris points out, "When the business expands beyond the point that the
owner can control everything himself, serious problems are encountered.
The ability to delegate responsibility and authority while still keeping
control, is generally lacking." The type of training required for over-
coming these difficulties might thus be somewhat different than that
required for ensuring the proliferation of new firms. In the latter case,
the existing apprenticeship system, the training ground for new entre-
preneurs in most countries, may play a more crucial role.
With respect to the other socioeconomic variables that might be
expected to affect the supply of rural entrepreneurship, there is little
empirical evidence. The few existing studies generally have not been
able to verify that sociological factors, such as caste, ethnicity, and
parent's occupation, are important determinants of entrepreneurial supply
(see Harris, 1970; Nafziger, 1971). Many of these socio-cultural variables
are interrelated with economic ones, and indeed interact with them.
Consequently, the individual effects cannot be easily measured.
Moreover, the various economic and socio-cultural determinants
of entrepreneurial supply most likely vary from country to country.
Thus detailed studies of entrepreneurial success would be useful
before designing policies and programs for rural non-farm enterprises.
The major policy and program issues will now be examined.
4. MAJOR POLICY AND PROJECT ISSUES
The foregoing analyses and other studies have established that the
rural non-farm sector is much more important than has been generally
recognized and, given even a neutral economic environment, the sector
could contribute much more to employment, equity, and output objectives.
Unfortunately, the economic policies of most developing countries are
generally biased in favor of large, mostly urban, capital-intensive
activities. The direct or indirect consequences of many of these policies
generally have been strongly negative for the non-farm sector, and
specific policies pertaining to the sector have been given little
An analysis of the policy environment within which rural non-farm
activities operate is consequently an important ingredient in the formu-
lation of specific projects and programs. Such an analysis would reveal
the degree to which the existing policies affect the viability of various
rural non-farm projects as well as indicate the potential efficacy of
various policy changes. Indeed, changes in existing policies might even
reduce or remove the need for specific projects in this area.
This policy analysis, however, must take account of not only
economic, but political considerations as well. The government's real
attitude toward rural non-farm activities is often difficult to ascertain
and frequently its position is not clear-cut or self-evident. The government
may assert strong support for rural non-farm activities in its Plan,
yet allocate few resources to this area, and may, in fact, be pursuing
a set of policies and programs for large-scale enterprises that have
a strong negative impact on the rural non-farm sector.
Since the national governments frequently consist of or represent
those individuals who benefit from the existing structure of the economy,
they may be reluctant to institute new policies and programs that might
adversely affect their interests. If these individuals are benefiting
from existing policies and programs favoring large-scale enterprises
they may be unwilling to institute even "neutral" policies for rural non-
farm enterprises whenever these are perceived to have potentially
deleterious effects on larger scale enterprises. Moreover, small rural
non-farm enterprises are widely dispersed and politically are frequently
neither well organized nor effectively represented. Presenting govern-
ments with an analysis of the effects of various policy and program
changes on existing interests, including a determination of the dis-
tribution of costs and benefits, is important to permit judgements as
to the government's willingness and ability to institute any policy
In this portion of the paper, specific issues will be raised with
respect to the major policy variables that influence non-farm activities.
Policies causing input price distortions will be discussed first, followed
by a brief discussion of the major nonprice distortions. A brief dis-
cussion of policy variables which influence demand for the sector's
output will conclude this section. Following this discussion of policy
issues, questions will be raised with respect to the major types of
projects designed to affect rural non-farm enterprises.
4.2 Policies Resulting in Factor Price Distortions
Distortions of input prices often originate from unsuspected
sources and, as emphasized in the earlier section, have very significant,
pervasive effects on rural non-farm activities. Five of the major sources
of price distortion will be discussed interest rates, tariff rates,
foreign exchange rates, fiscal policies, and minimum wages.
4.2.1 Interest Rates
In most developing countries, two distinct capital markets exist -
the "formal" and the "informal." Banks and similar institutions con-
stitute the formal market while money lenders, raw material suppliers,
and purchasers constitute the bulk of the informal market. Interest
rates vary widely between the two. For eight countries shown in table
4.1, official interest rates, where government imposed ceilings
generally exist, ranged from 9 to 24 percent, while the "nonofficial"
rates ranged from 29 to over 200 percent. Particularly under inflationary
conditions, the formal real rates become very low, sometimes negative.
Thus, not surprisingly, the banks have tended to lend only to the es-
tablished, large-scale firms, which may appear to the banks to involve
lower risks and lending costs. Most of the recipients are urban based
and they have tended to become more capital-intensive than would have
been the case at the "opportunity costs" of capital. For the rural non-
farm activities, an important question is: To what extent has the frag-
mented capital market resulted in depressed enterprise creation, capital
formation, employment generation, and labor productivity?
A COMPARISON OF OFFICIAL AND NONOFFICIAL RATES OF
INTEREST IN SOME DEVELOPING COUNTRIES
Country Rate of Interest Rate of Interest
Thailand 29% 9.0%
Korea 35-60 17.5
Afghanistan 33 ---
Gambia, Sudan, 50-100 10-12
and Sierra Leone
Colombia 36-60 24.0
Haiti 40-240 12-15
1. Thailand: Ingle, et al. (1973)
2. Korea: Morrow and White (1973)
3. Afghanistan: Norvell (1973)
4. Gambia, Sudan: United Nations, F.A.O. (1963)
Sierra Leone: Bank of Sierra Leone (1969)
5. Colombia: Kochav, et al. (1974)
6. Haiti: Haggblade, Defay, Pitman (1979)
An important related issue is the extent to which the government-
imposed ceiling on "formal" interest rates contributes to the gap
between the demand for and supply of credit for rural non-farm enterprises
at a given time in a country.1 An artificially depressed rate of interest
can be expected to reduce credit availability in rural areas as compared
to what would prevail if the rates were equated with "opportunity costs."
The import duty structure can be an important source of differential
treatment for the urban large-scale over the rural small-scale enterprise.
For most developing countries, import duties are lowest for heavy capital
goods and become progressively higher through intermediate and consumer
durable goods categories (see table 4.2). Yet, many items classified as
intermediate or consumer goods in tariff schedules are capital goods
for rural small-scale firms. In Sierra Leone the sewing machine, an
important capital item for tailoring firms, was classified as a luxury
consumer good and taxed accordingly (Liedholm and Chuta, 1976).
Further escalating the distortion in capital cost is the frequent
practice of granting concessions or even total waiver of import duties
on capital goods or raw materials for specified periods as an inducement
for industrial development. In some cases, small firms may technically
qualify for similar concessions, but may be unaware of this opportunity
or, even when aware, they may find the process so complicated and time
See below, p. 66, for a more detailed discussion of credit programs.
AVERAGE TARIFF RATES BY END-USE GROUPS FOR
SOME DEVELOPING COUNTRIES
Group Ghana Pakistan Brazil
1966 1965/66 1964
I. Capital goods:
II. Intermediate goods:
1. For industries
2. For industries
III. Consumer durables
IV. Consumer nondurables
Ghana: Steel (1972, p. 219)
Pakistan: Lewis (1970, p. 68)
Brazil: Bell (1971, p. 47)
consuming that it is not economic for them to exercise the option. In
many other cases, small firms do not even qualify.
4.2.3 Foreign Exchange
Many developing countries institutionally maintain a high price
for foreign exchange but grant concessionary rates to large firms.
Small firms are deprived of comparable advantages since they usually
do not qualify for concessional rates. Even if there are no concessions
the large firms usually import relatively more equipment and inputs
and therefore benefit more than the small ones. The consequence is
to encourage greater capital intensity among urban large-scale industries
and a less than optimum use of capital among rural non-farm industries.
Under conditions of rationed foreign exchange, the larger firms are more
likely to have access to these scarce resources than the small firms.
4.2.4 Other Tax Incentives
Several countries employ tax incentives to encourage industrial
development. These incentives have differed with respect to timing
and coverage. Qualifying firms have been eligible for such incentives
as: (1) tax holiday periods, (2) accelerated depreciation and invest-
ment allowances, and (3) exemptions from some import duties, as discussed
above. These incentives provide an advantage to those enterprises able
Many of these fiscal incentives have had pronounced differential
effects as between large and small rural non-farm firms. Income
Tariffs can also bring direct benefits to rural non-farm enterprises
when they are placed on commodities that the small enterprises produce.
These may benefit large, local enterprises as well.
tax exemptions in many countries are only made available to enterprises
above a certain minimum investment or employment threshold. In those
countries with no minimum requirements, the qualifying procedures are often
so sophisticated and time consuming that they discourage small entrepre-
neurs. On the other hand, smaller enterprises, particularly those at
the lower end of the size spectrum, often do not pay any income taxes
due either to explicit size exemptions or to a tax administration that
is ineffective in reaching the smallest firms.
There is also an array of indirect taxes that may discriminate
against smaller enterprises. Sales and excise taxes, which apply at
all stages in the production process fall more heavily on smaller
enterprises, which usually must pay taxes on each input, than on the
larger, vertically-integrated enterprises (World Bank, 1973). Such
taxes as licensing and municipal fees and stamp taxes tend to be re-
gressive in nature.
4.2.5 Minimum Wage Regulations
Minimum wages, often initiated to achieve socially sound objectives,
often apply only to larger enterprises in urban areas of developing
countries. Where they are applicable country-wide, they are often not
enforced as effectively among the smaller-scale, rural non-farm
activities. Minimum wages in most developing countries have tended to
cause greater capital intensity in urban areas and a greater rural-
to-urban migration. The latter may be the consequences of potential
migrants' perception rather than actual urban employment opportunities
and thus exacerbate the (perhaps partially revolving) urban unemployment
problem. For the rural areas, the overall direct effect has been
a possible reduction in the number of potential entrepreneurs (Berg,
1966, p. 201) and a deterrent to development of a permanent skilled
rural labor force (Weeks, 1971; Berg, 1966; Isaac, 1963).
4.3 Policies With Nonprice Supply Effects
Some policies employed by developing country governments have effects
on rural non-farm activities other than through input prices. These may
originate from unexpected sources and can range from very general to
rather specific influences. These are discussed under two categories -
infrastructure and industrial policies.
4.3.1 Development of Infrastructure
Policies designed to develop the infrastructure of a developing
economy could indirectly affect the performance of rural non-farm
enterprises. The provision or expansion of electricity, water, or roads
would appear to benefit these enterprises (World Bank, 1978a). These
same amenities also benefit their larger scale urban-based counterparts,
which may now be able to enter markets previously dominated by rural
non-farm enterprises. Indeed, one differential advantage possessed by
rural non-farm enterprises may be that they do not require large amounts
of potentially costly infrastructure.
If clear-cut advantages of expanded infrastructure are indicated,
two guidelines are relevant: (1) the substantive nature and/or sophis-
tication of the installation or service, and (2) the scale or rate of
installation of the service. In the case of electricity, for example,
the initial installations may be small on-premise generating units
for key enterprises; next there could be larger generators serving
several enterprises, followed by phasing up to community or area-wide
service from a single generator. An important overall objective is
the compatability between the infrastructure contemplated and the
enterprise clientele to be serviced, if there is to be balance between
social costs and benefits. It is also important to make sure that
rural non-farm enterprises are not discriminated against in their
access to the infrastructure once it has become available.
4.3.2 Industrial Policies
Many policies designed primarily with reference to large-scale
urban firms are also made applicable to small firms, but often work
a relative hardship on the latter. Simple licenses or permits to engage
in business may be so exacting in requirements or so formidable admin-
istratively as to constitute virtual barriers to entry for small firms.
On the other hand, some countries, such as India, have taken positive
measures by reserving certain business activities for the rural non-farm
or small-scale sector. Such actions pose an issue as to whether they
may bring over-corrections and raise barriers to development of other
Conditions of employment and product standards specified by govern-
ment can be obstacles for the small firms if the measures do not take
into account sufficiently the realities facing such firms. These
regulations can have both positive and negative impacts on consumers
as well as on the firm and its workers. Meritorious are those which
realistically safeguard the health and safety of workers and the
Quantity controls of different kinds also can have relative or
absolute negative effects on rural non-farm activities. Examples are
import quotas on raw materials, such as special cloth for small textile
producers. In any general quota program, the large-scale firms are
usually more likely to be accomodated than are the small ones.
4.4 Policies Affecting the Demand for Rural Non-farm Activities
The analyses in the earlier sections of this paper have revealed
that the primary demand for most rural non-farm goods and services
stems from the agricultural sector and that this demand is transmitted
through both income and production linkages. Since the available
evidence indicates that the rural households' income elasticity of
demand for rural non-farm goods is positive and agriculture generates
the largest share of rural income, policies designed to increase agri-
cultural output and/or income have an important indirect effect on the
demand for rural non-farm activities. Consequently, pricing policies
that improve the terms of trade between agricultural and the large-
scale urban sector or specific investment programs and policies designed
to increase, directly or indirectly, agricultural production and income
can generate an increased demand for a wide array of rural non-farm goods
The nature and composition of these agricultural policies and pro-
grams, however, should also be considered, since they can have important,
differential effects on the demand for rural non-farm activities. There
Demand is also crucially affected by specific marketing projects,
such as government purchasing programs or export marketing schemes.
These and others are described below on p. 76.
is some evidence that the higher income rural households have a somewhat
lower income elasticity of demand for rural non-farm activities than do
the lower income households, the majority of which are small-scale
farmers (see King and Byerlee, 1978). Moreover, as noted previously,
the agricultural inputs such as tractors and fertilizers used by large-
scale, high income farmers are less likely to be produced in rural
localities than are those inputs used by the small-scale farmers (see
Johnston and Kilby, 1975). Consequently, policies and programs designed
to benefit a larger number of small-scale, low income farmers are likely
to generate a larger demand for rural non-farm activities and services
than those designed to benefit a few, larger scale farmers. These
differential effects on rural non-farm activities must be recognized
when designing agricultural policies.
Government policies also can affect the demand for rural non-farm
activities that arise from production linkages with large-scale industry.
Sub-contracting is the most frequently discussed industrial linkage (see
p. 27). Properly designed, such policies can provide relatively stable
demand for certain products at prices which will not adversely affect
the profitability of rural enterprises or the quality of the work environ-
ment. Unfortunately, some policies of developing countries have had a
deleterious effect on sub-contracting. The system of "cascading" sales
tax, for example, whereby taxes apply to all stages of activities relating
The small scale farmers are also more likley to use primarily the
smaller, rurally-based agricultural processing establishments, while the
large-scale farmers might be expected to make more use of the larger scale,
urban-based processing plants.
to any one product discourage sub-contracting and encourages vertical
integration into single, capital-intensive units. Some of the previously
mentioned factor price distortions also tend to reduce the incentive
for large scale units to sub-contract with small, non-farm units.
4.5 Rural Non-farm Enterprise Project Issues
In most developing countries, important issues relating to those
existing or potential projects that directly influence rural non-farm
enterprises also need careful examination. Some of the issues are of
a general nature while others are project specific, but all are important
for consideration in the process of designing, implementing, and
evaluating rural non-farm enterprise projects.
4.5.2 General Project Issues
One of the major general issues confronting designers of projects
for rural non-farm enterprises is: How does one identify the intended
project beneficiaries? The descriptive profile presented earlier in
this paper has revealed that, although most of those engaged in rural
non-farm activities are relatively poor, the enterprises themselves are
heterogeneous and widely dispersed geographically. Since it may be
impossible with a single project or set of projects to address the
needs of all rural non-farm enterprises, it will usually be necessary
to develop some criteria for more sharply delineating the project
beneficiaries. Geography, enterprise size, or enterprise types are
some elements that might be incorporated into the formulation of
1Specific programs and projects designed to facilitate sub-contracting
such as information dissemination, quality control, management training, are
described in Staley and Morse, 1965, chapter 9. The minimum wage policy
might tend to encourage sub-contracting while capital subsidization and
other policies would tend to discourage it.
these criteria. It would be useful to identify the types of rural
non-farm activities and the specific enterprises within them that
possess the greatest potential for expanding employment, output and income
in the rural areas. Such information, however, is not generally available
and some in-depth surveys may need to be undertaken in the early stages
of project preparation.
A related general issue confronting designers of rural non-farm
enterprise projects is: How does one decide the types of direct assistance
to be provided? A crucial element involves the identification of the
types of constraints facing the rural enterprises. Only after these
constraints have been identified, can the type and nature of direct
assistance be ascertained. Otherwise, inappropriate interventions may
A final general issue is: How does one establish an effective
monitoring and evaluation system for these projects? Some useful ex-
post evaluations of rural non-farm projects have been undertaken (see Kilby,
1979). Yet, a crucial element that is frequently missing is adequate
benchmark data from which performance can be measured. Once again,
the need for detailed surveys in the project preparation stages becomes
4.5.3 Specific Project Issues
Various types of direct assistance have been utilized in developing
countries to promote rural non-farm enterprise. Although there does
not appear to be any standard way of classifying the various kinds of
direct assistance (see, for example, World Bank, 1978a, 1978b; Staley and
Morse, 1965), a listing of the types of direct intervention would include
the provision of credit, technical/production, management, and marketing
assistance and common facilities (usually industrial estates). The
form which each type of assistance takes and its associated delivery
channels are presented in table 4.3. The major issues that relate to
each of these types of direct assistance for rural non-farm enterprises
will now be examined.
18.104.22.168 Credit Assistance
There are several issues related to the design of credit projects
for rural non-farm enterprises. These issues can be usefully grouped
into those that center on the rural non-farm enterprises' demand for
credit and those that relate to the supply of credit to these enterprises.
One important demand issue is: What is the extent of the effective
demand for credit by rural non-farm enterprises? Some evidence would
appear to indicate that this demand is quite sizable. Rural non-farm
entrepreneurs, for example, when asked directly to identfiy their
greatest assistance needs and greatest perceived bottleneck, will usually
list credit and capital first (Liedholm and Chuta, 1976; Schatz, 1977;
Haggblade, et al., 1979). There is evidence that for many types of rural
non-farm enterprises, the rates of return on existing capital are quite
substantial (see above, p. 46). These high rates of return indicate
that the potential demand for credit could be quite large.
Yet, other evidence indicates that the rural non-farm enterprises'
demand for credit may be somewhat less extensive than indicated above.
Detailed analyses of entrepreneurs undertaken in Sierra Leone (Liedholm
and Chuta, 1976) and Kenya (Harper, 1978) revealed that, although lack
of credit was perceived to be the crucial bottleneck, some other problems,
TYPES OF ASSISTANCE, THEIR VARIOUS FORMS AND DELIVERY CHANNELS
Assistance Form Delivery Channel
-Loans in cash and/or kind
for fixed assets and/or
-Advice on processes, design
of products, tools, equipment,
machines, quality control,
-Capital budgets, etc.
-Advice on packaging,
merchandising, product demand
-Raw material procurement
-Maintain emporia sales and
displays at home and abroad
-Serve as collection centers
-Buy on consignment basis
-Undertake export service
-Offer credit insurance
-Electricity and water
-Private Voluntary Agencies
-Vocational Training Institutes
-Extension on-the-spot, at
Development Centers or
through mobile workshops
-Appropriate technology units
-Vocational Training Institutes
-Extension on-the-spot, at
the Industrial Development
Centers, through mobile
-Formal and informal meetings
-Credit and Export Schemes
-Customer Service Centers
-Industrial Estates, Areas
such as inadequate management or raw material procurement difficulties,
proved, in actuality, to be the crucial basic constraints facing these
enterprises. Unless these other difficulties are recognized and dealt
with, the simple provision of credit could, at a minimum, be wasteful
and could actually prove to have a deleterious effect on the rural non-
farm enterprises. Several studies have indicated that at least some
of the demand for finance is met through self-financing or through
the informal credit market. Recent evidence from Sierra Leone (Liedholm
and Chuta, 1976), Ghana (Steel, 1977), Bangladesh (Ahmed, et al., 1978),
and Haiti (Haggblade, et al., 1979) reveal that personal and family savings
accounted for over 80 percent of the funds used for establishing rural
non-farm enterprises, while about 90 percent of the funds used for ex-
pansion were reinvested profits. Are these self-financing and informal
credit sources sufficient to satisfy the major capital needs of these
enterprises? One might argue that rural non-farm enterprises are
simply forced to use these sources because they are denied access to
the apparently lower cost, "formal" credit markets. Clearly, more
studies are needed to ascertain the magnitude of the rural non-farm
enterprises' effective demand for credit.
Another demand issue relates to the composition of this credit
demand from rural non-farm enterprises. In particular, is the credit
demand primarily for fixed or for working capital? The composition of
the credit demand does appear to vary somewhat depending on the size and
1This is a supply issue, which will be examined below. Although
the interest rate in the formal market may be low, the borrowing costs
may be high due to high transactions costs. The complicated procedures
and long processing time required by these institutions may make the
"transactions" costs very high if not prohibitive, particularly for
those enterprises needing working capital quickly.
type of rural non-farm enterprise. For the smallest enterprises, which
account for the bulk of the rural non-farm sector, the primary credit
demand appears to be for working capital (see Haggblade, et al., 1979).
It is important to ascertain how much of the credit demand is simply a
manifestation of some other problem such as a raw material shortage,
inadequate management or a lack of demand.
One important supply issue centers on what is the appropriate
delivery channel for providing financial services to rural non-farm
enterprises? In most developing countries, formal credit institutions
such as commercial banks, specialized small enterprise banks, specialized
divisions of development banks, credit unions, cooperative and worker
banks have typically been used to channel funds to rural non-farm enter-
prises. Although such devices as rediscounting facilities, guarantees,
and earmarked funds are frequently introduced to entice these "formal"
institutions into expanding their lending to rural non-farm enterprises,
it is not yet clear that these have been successful in significantly
expanding the amount of formal credit available to these enterprises,
particularly the smaller ones.1
Indeed, the vast majority of these rural non-farm enterprises have
never even applied for funds from formal credit institutions. Thus,
alternative institutional mechanisms to the formal ones might also need
to be considered. Informal financial institutions such as money lenders,
input suppliers, purchasers and rotating credit societies, provide most
For a review of alternative institutional arrangements, see Kochav,
et al., 1974.
21n Haiti, 94 percent had never applied (Haggblade, et al., 1979),
while in Sierra Leone, the figure was 96 percent (Liedholm and Chuta, 1976).
of the institutional credit to rural non-farm enterprises. Studies
of these institutions might reveal how these formal and informal
financial institutions are linked and how the informal institutions
might be better integrated into the formal credit system. Finally,
consideration might also be given to establishing new intermediary
institutions, possibly linked with private voluntary organizations,
which might, in turn, link into the formal financial system. Clearly,
no single delivery system has emerged as a solution for providing credit
to the wide array of rural non-farm enterprises.
Another related supply issue centers on the costs and risks associated
with lending to rural non-farm enterprises. It is argued that, owing to
the geographical dispersion and vast number of rural non-farm enterprise-
borrowers, the administrative costs of lending to this group are signif-
icantly higher than lending to large-scale borrowers. Another allegation
which is often made against lending to rural non-farm enterprises is
that, due to the vulnerability of these firms to adverse economic con-
ditions, failure rates are high and therefore the incidence of default
is higher compared to large firms (World Bank 1978b). Furthermore,
the risk factor in lending to rural non-farm enterprises is compounded
by the fact that these firms possess very little collateral and the actual
losses of principal could be fairly high.
These contentions, while valid in some cases, need to be tempered
somewhat. There is evidence in some countries that the default rates
In the Philippines, it was estimated that the administrative costs
incurred by two lending institutions for small-scale loans were 3 percent,
for a given value of lending, while those for large-scale loans were 0.5
among small farmers and small businessmen are lower than among larger
borrowers (Steel, 1976, p. 182; Meyer, 1978). Moreover, alternative
procedures for reducing risks and administrative costs can perhaps be
developed. Instead of relying only on collateral, which many rural
non-farm enterprises may lack, the underlying economic viability and
cash flow potential of the borrowers can be emphasized. Also, loan
repayment schedules can be designed to fit the nature of the different
rural non-farm enterprises. The appraisal and supervision of the loans
for rural non-farm enterprises could be standardized and streamlined.
In addition, managerial, technical and other assistance could be pro-
Vided both borrowers and lenders. Finally, if loans are made at rates
which even approximate the opportunity cost of capital, the financial
institutions could probably cover any higher costs associated with
lending to rural non-farm enterprises.
A final issue that relates both to the demand for and the supply
of credit is: What should be the interest rate for small-scale enterprise
credit projects? In many countries, and within some donor agencies,
there is a feeling that rural non-farm enterprises should receive credit
at a rate below the opportunity cost of capital. It is argued that these
enterprises generally are owned by the poor and that they are already
operating in a policy environment that discriminates against them. More-
over, they contend, if large-scale enterprises already receive subsidized
interest rates, then it would be unfair, and politically unwise, to charge
higher interest rates to rural non-farm enterprises.
A related issue is whether or not the credit and noncredit assistance
should be separated or provided by the same institution.
There are persuasive arguments for an interest rate that closely
approximates the opportunity cost of capital. First, a subsidized
interest rate may encourage rural non-farm enterprise to adopt tech-
nologies that are too capital-intensive. Secondly, most studies (see
Liedholm and Chuta, 1976; Haggblade, et al., 1979) indicate that the
viable rural non-farm enterprises are able and willing to pay higher
interest rates than currently exist in the formal market, especially
if the only alternative is the informal money lender's very high rate.
Higher interest rates may also serve to increase the rate of savings
in rural areas. Finally, the economic viability of the lending insti-
tution is seriously compromised if the interest rate is unduly sub-
sidized. Yet, although strong arguments exist for interest rates closely
approximating the opportunity cost of capital, political, social, and
economic realities may necessitate some degree of interest rate sub-
sidization for rural non-farm enterprise projects.
22.214.171.124 Technical Assistance
Several major issues need to be considered while designing technical
assistance for rural non-farm enterprises. First, it is important to
ascertain the magnitude of the demand for technical assistance. There
is some evidence to indicate that entrepreneurs are not aware of their
.need for technical assistance and the benefits they may derive from it
(see Liedholm and Chuta, 1976; Harper, 1978). Consequently, case studies
may be needed to identify the amounts and required forms of technical
assistance, particularly since assistance needs will likely vary by the
size and type of enterprise.
A second general issue which deserves careful consideration is:
What are the most cost-effective institutional mechanisms, if any, for
delivering technical assistance to rural non-farm enterprises? One
important channel that has been utilized in some developing countries
such as Kenya (Harper, 1978), Ghana, and India is a rural industry
extension service. Unfortunately, there are no systematic analyses
of the experiences with or the effectiveness of this approach. An
important issue which needs careful consideration is: What quality
of personnel should be utilized for delivering these services to rural
non-farm enterprises? Is there a place for both generalists and
specialists in view of the heterogeneity of rural non-farm enterprises
and the complex range of technical operations they face? Another
related issue is whether these services should be offered free, at some
token fee, or at economic costs to be paid by the rural non-farm entre-
preneurs. If extension services should be delivered free to rural
non-farm enterprises, who should underwrite the necessary costs and
for how long? A final issue is whether this active outreach approach
is as cost effective as a more centralized assistance approach, in which
clients would approach a centrally located institute for assistance.
The Rural Industrial Centers in Kenya (see Livingston, 1977) and the
Industrial Development Centers in Nigeria (see Hawbaker and Turner, 1972)
are examples of the more centralized approach to the delivery of assistance.
Vocational training institutes also have been relied upon for
delivering technical assistance, particularly to unemployed youth in
developing countries. For some already employed personnel, vocational
training has been utilized for developing alternative job opportunities
or enhancing capabilities in existing lines of activity. But, unfor-
tunately, the progress of clientele has not been ascertained and overall
effectiveness of vocational training has not been determined.
An already existing institutional mechanism is the apprenticeship
system, and an important issue is whether or not it can be effectively
utilized for transferring technical skills to existing and potential
rural non-farm enterprises. The available empirical evidence reveals
that in most developing countries, the apprenticeship system is the
primary vehicle for skill formation in rural non-farm industry (Liedholm
and Chuta, 1976; Steel, 1977; Haggblade, et al., 1979). Recently,
Mabawonku's study (1979) revealed that with respect to both private and
social rates of return in employment, apprentice training compares
quite favorably with government trade and vocational school programs.
Yet, the apprenticeship system, which is an important channel for skill
formation in rural non-farm industry, is often overlooked by donor
agencies and host governments of developing countries when designing
technical assistance programs.
Another related issue is the degree to which technical assistance
delivery institutions should be separated from the regular governmental
machinery. There would appear to be some advantages associated with
utilizing semi-public organizations or even private sector approaches;
confidence among entrepreneurs would likely be higher and qualified
staff could perhaps be more easily recruited and retained. At the same
time it is unlikely that, in practice, donor agencies can remain com-
pletely detached from the existing governmental technical assistance
Finally, there is the question of whether there are effectively
staffed in-country research institutions to address relevant researchable
problems of small rural non-farm enterprises. Are the institutions
linked to comparable research centers in the world and, within the
country, are there effective links with personnel handling information
dissemination and technical advisory services?
126.96.36.199 Management Assistance
An important issue which should be addressed while designing a
rural non-farm project is to ascertain whether or not there is a demand
for management assistance. Indeed, most rural non-farm enterprises
scarcely recognize that lack of management capacity could pose a serious
problem. Yet, previous studies of rural non-farm enterprises have
revealed that managerial competence is a key determinant of business
success (Harris, 1967; Liedholm and Chuta, 1976; Steel, 1977). It is,
therefore, crucial to ascertain not just what are the perceived, but
also the actual needs in this area.
Another related issue which must be considered is what forms of
management skills these rural non-farm enterprises really need; these
needs will likely vary somewhat depending on the size and nature of the
enterprises. Most rural non-farm entrepreneurs use simple technologies
and possess little or no educational background. Consequently, it may
be important to design management assistance to enable rural non-farm
entrepreneurs to be able to distinguish between personal and business
transactions, evaluate resources (especially their labor input) at their
appropriate opportunity costs, understand effective inventory plans,
adjust their businesses to viable sizes, and adopt methods of pro-
duction that utilize local resources.
A third issue which deserves careful consideration is what delivery
mechanism, if any, will be cost-effective in carrying out management
assistance projects for rural non-farm enterprises. This issue, which
raises other questions relating to the appropriate use of an apprentice-
ship system, extension services and other institutional mechanisms,
has already been discussed under issues relating to technical assistance.
188.8.131.52 Marketing Assistance
The design of marketing assistance for rural non-farm enterprises
raises several issues. First, it is important to ascertain what existing
or new sources of domestic demand are available and how these could be
further stimulated or developed. Many of these sources and programs
have been discussed previously. In addition, the governments themselves
have frequently developed programs to purchase the products of rural
non-farm enterprises, but sales to governments have been hampered by
cumbersome purchase procedures and unrealistic quality standards
(Schatz, 1977,p. 199). A relevant issue is whether official procedures
should be streamlined and excessively high quality requirements waived
to facilitate the purchase of rural non-farm enterprise products by
government departments and the public?1
Large-scale enterprises frequently use packaging or promotional
devices to create product "quality" differentials. Some of these "quality
differentials" may be specious and can have a deleterious effect on rural
non-farm enterprises operating in these same industries. The Kenyan soap
industry (Langdon, 1975) and the Egyptian carpet industry (El Karanshawy,
1975) provide illustrative examples.
A second issue that is relevant for designing marketing assistance
relates to the external demand for the products of rural non-farm
enterprises. In particular, a key issue is how one develops and delivers
information to rural enterprises on the details of both the existing and
new product demand in foreign markets as well as information on product
handling and financial transactions. In addition, since products of
rural non-farm enterprises must be competitive in foreign markets, a
related issue is how one ensures at least minimum product quality.
Thirdly, there is the issue of whether there exists an accessible,
cost-effective, institutional support which can enable rural non-farm
enterprises to purchase raw materials and produce for and effectively
reach the export markets. In most countries this institutional support
is urban based. Decentralization of such facilities to service the needs
of rural non-farm enterprises becomes crucial.
With respect to raw material purchase, evidence from some developing
countries such as Haiti and Bangladesh (Haggblade, et al., 1979; Ahmed,
et al., 1978) reveals that the lack of raw materials constitutes
a major constraint in rural non-farm enterprise. An important issue
which deserves careful attention is what forms of delivery channels
will be cost-effective in providing raw materials to rural non-farm
enterprises? Some developing countries' governments have relied on the
formation of rural cooperatives or producer associations for bulk purchasing
of inputs in order to lower costs of production. However, evidence from
some developing countries reveals that rural cooperatives have often failed
due to personal rivalry, lack of effective leadership and management
problems (Shetty, 1963, pp. 184-185). In other countries, such as Honduras,
the government operates a raw materials bank so that scarce intermediate
inputs can be provided to rural non-farm enterprises; administrative
procedures for distributing inputs often become quite cumbersome.
184.108.40.206 Common Facilities
In many developing countries, the most popular type of assistance
used in providing common facilities for rural non-farm enterprises is
industrial estates. In some developing countries, industrial estates
have been utilized for decentralizing industry toward small rural towns
and villages (Dhar and Lydall, 1961, p. 36). An important issue which
arises is whether estates located in rural areas where basic infra-
structural facilities are lacking can be cost-effective. Experience from
India, reveals that there are relatively few economic justifications
for establishing estates in rural areas (Kochav, et al,. 1974, p. 33).
Rural non-farm activities in developing countries have begun to
receive increased attention from international assistance agencies and
the governments of developing countries. This growing interest has
paralleled the increased international concern for equity and employment
objectives and the realization that expanded rural non-farm activities
might contribute to both growth and improved equity within countries.
Unfortunately, there have been relatively few empirical studies of these
activities. Consequently, those charged with formulating and executing
programs and policies to expand productive rural non-farm employment
and earning opportunities have been generally forced, of necessity, to
make decisions "unencumbered by information." In order to fill that
void, this paper has attempted to assemble and interpret the currently
existing data concerning rural non-farm activities. The major issues
relating to these activities have been considered in the body of the
paper and are summarized briefly below.
A first major issue concerns the overall importance and composition
of rural non-farm activities. The available evidence indicates that
rural non-farm activities are quantitatively very important with from
30 to 50 percent of the rural labor force in most developing countries
either primarily or secondarily engaged in some form of rural non-farm
activity. Currently, the rural non-farm sector encompasses a wide
variety of activities, although manufacturing, commerce, and services
generally predominate. Manufacturing appears to be particularly significant;
in fact, employment in rural manufacturing often exceeds that in urban
Related to the issue of the current importance of rural non-farm
activities is the question of how rural non-farm activities have evolved
as development proceeds. Hymer and Resnick, for example, hypothesize
that these activities have declined and will continue to do so as rural
incomes rise and opportunities for trade increase. Although some
specific types of rural non-farm activities appear to have declined
over the recent past, the empirical evidence indicates that, overall,
rural non-farm activities and employment have been increasing in most
Several important issues, which have additional implications for
the future growth of the sector, relate to the nature of the demand for
the goods and services produced by rural non-farm activities. One crucial
issue, on which there has been a divergence of opinion, is whether or
not the demand for these activities increases as rural incomes increase.
Hymer and Resnick have argued that rural non-farm activities are "inferior,"
which means that the demand for them would be expected to decline as
rural incomes rise. Mellor, Liedholm and Chuta, and various I.L.O.
Missions, on the other hand, have contended that there is a strong,
positive relationship between income and the demand for these activities.
The few empirical studies of rural demand, particularly that of King
and Byerlee, support the latter position.
Another demand-related issue is whether or not there are strong
backward and forward linkages between rural non-farm activities and
other sectors of the economy, particularly agriculture. Hirschman has
contended that linkages between agriculture and other sectors are quite
weak, while others, such as Mellor and Johnston and Kilby, have argued
that the linkages between rural non-farm activities and agriculture, in
particular, are or could be potentially very strong. The available
empirical evidence indicates that these linkages are quite important. The
rural non-farm sector is influenced by the pattern of agricultural growth,
but, also, the rural non-farm sector can influence the course and rate of
agricultural development. The evidence on linkages with large-scale
industry is sparse but the evidence available indicates that they are
somewhat limited. Finally, there is some empirical and analytical evidence
that the international market is an important component of demand for
certain types of rural non-farm products.
With respect to supply, one important issue is whether or not
rural non-farm activities are more labor-intensive and thus generate
more employment per unit of capital than other non-farm components of
the economy. The available empirical evidence is generally quite con-
sistent in indicating that small-scale, rural enterprises are more labor-
intensive than their larger-scale counterparts.
A key related issue is whether or not these same labor-intensive
rural non-farm enterprises use the scarce factor, capital, more efficiently
than other larger-scale enterprises. Several international groups and
individuals, including Nicholas Kaldor, have argued that the capital
productivity (i.e., the output-capital ratio) of small, rural enterprises
is lower than that of their larger-scale counterparts. Marsden, Liedholm
and Chuta and others have contended that the reverse situation holds.
The available aggregate country data are generally not of high enough
quality to provide a conclusive answer to these conflicting views,
although there are many instances where the small, rural non-farm
enterprises appear to possess the higher capital productivity. When
rural non-farm and urban large-scale enterprises within the same narrowly-
defined industry are compared, there is evidence, in several industries,
that the rural non-farm enterprises are not only more labor-intensive,
but also more productive per unit of scarce capital than their larger-
scale counterparts. Consequently, in these cases, there need not be a
trade-off between output and employment objectives, at least in a static
These findings are reinforced by evidence on profitability, which
indicates that profit rates in many rural non-farm enterprises are also
higher than those in urban, larger-scale firms. Moreover, in a dynamic
context, there is no empirical evidence to support Galenson and
Leibenstein's contention that the profit, savings, and reinvestment
rates of small-scale or rural non-farm enterprises are necessarily lower
than those of the large, capital-intensive enterprises.
Although static and dynamic efficiency considerations are of great
importance to policymakers, the equity implications of rural non-farm
activities also are attracting increased attention. The limited evidence
shows that, on the average, the income of rural non-farm households
is somewhat higher than that of farming households, but is substantially
below urban incomes. Rural non-farm activities are generally undertaken
by very small-scale, artisan and informal enterprises, which employ on
the average, fewer than five individuals. These activities are partic-
ularly important for those rural households with little or no land.
Given the great potential of rural non-farm activities for increased
employment, increased income, and favorably affecting income distribution,
many governments are showing increasing interest in assisting rural
non-farm enterprises. Governments can assist these enterprises by
general policy measures, which affect the environment in which rural
non-farm enterprises operate, and by providing direct project assistance.
Several major policy options are available to those governments
interested in influencing rural non-farm activities. However, great
care must be exercised in policy selections as many government actions,
seemingly unrelated to rural non-farm activities, can have inadvertently
adverse effects on them. For example, policies that result in input
price distortions have significant, though often unintended, negative
effects on rural non-farm activities. In most developing countries,
interest rates, tariff rates, foreign exchange rates, and tax policies
have been designed to benefit large-scale enterprises and consequently
are generally biased against the small, rural non-farm enterprise.
Government policies with respect to the infrastructure, industry, and
agriculture also have important indirect effects on the expansion of
rural non-farm employment and income opportunities. Because of the
strong linkages between agricultural and rural non-farm activities,
agricultural policies and programs, in particular, have a strong
influence on rural non-farm activities.
The major types of direct assistance projects used to promote
rural non-farm activities include a broad spectrum of interventions:
the provision of credit, technical, management, and marketing assistance
and common facilities (usually industrial estates). A crucial element
in determining which form of direct intervention is most appropriate
is the identification of the key constraint or constraints facing the
rural enterprise. Rural entrepreneurs, when asked to identify their
primary constraint, will usually state that it is a lack of credit.
Yet, in-depth analyses frequently reveal that other underlying constraints
are more crucial. Such analyses are important for identifying both
the types and forms of assistance that are most needed.
Another key project issue that deserves careful consideration relates
to the supply of this assistance. What are the most cost effective
mechanisms for delivering this assistance? Should existing institutions
be used or should new ones be established? Should these be separated
from the existing governmental structure? Should several forms of
assistance be provided by the same institution? Because rural non-farm
enterprises are so heterogeneous and conditions vary so markedly from
country to country, there are no single answers to these questions;
rather, the appropriate institutional arrangements will depend importantly
on the country and the nature of the non-farm enterprises that are to
receive this assistance.
Although rural non-farm activities represent only one facet of the
rural development process, their importance in this process is becoming
increasingly recognized. With judicious governmental policies and
carefully formulated direct assistance measures, the already sizable
contribution of rural non-farm activities to this process can be sig-
Ahmed, S.; E. Chuta, and M. Rahman. 1978. "Rural Industries in Bangladesh:
Research and Policy Implications," Journal of Management. (University
of Dacca) (April):25-37.
Anderson, Dennis, 1979. "The Financing of Small and Medium Enterprise in
the Philippines," Washington, D.C.: World Bank. (Mimeographed).
Baer and Herve, 1966. "Employment and Industrialization in Developing
Countries." Quarterly Journal of Economics (80):88-107.
Bangladesh Institute of Development Studies, 1979.
Project Phase I Report." Dacca, Bangladesh.
"Rural Industries Study
Bank of Sierra Leone, 1969. "Notes and Comments," Economic Review. 22:1-5.
Barber, Clarence L. 1969. "The Capital-Labor Ratio in Under-developed
Areas." Philippine Economic Journal. 8:85-89.
Baron, C. G. 1975.
"Sugar Processing Techniques in India."
Technology and Employment in Industry:
Geneva: International Labour Office.
in A. S.
Bell, Clive and Peter Hazell. 1976. "Preliminary Regional Analysis of
the Muda Irrigation Scheme." Washington, D.C.: World Bank. (Mimeographed).
Bell, Harry H. 1971. Tariff Profiles in Latin America. New York: Praeger
Berg, Elliot J. 1966. "Major Issues of Wage Policy in Africa." In
Arthur M. Ross (ed.). Industrialization and Economic Development.
Berna, J. J.
1960. Industrial Entrepreneurship in Madras State, India.
Asia Publishing House.
Bhalla, A. S. 1964. "Investment Allocation and Technological Choice A
Case of Cotton Spinning Techniques." Economic Journal. 74:611-622.
S1965. "Choosing Techniques: Handpounding vs. Machine Milling
of Rice: An Indian Case." Oxford Economic Papers. 17:147-157.
1975. Technology and Employment in Industry. Geneva: Inter-
national Labour Office.
Boon, G. K. 1975. "Technological Choice in Metalworking with Special
Reference to Mexico," in A. S. Bhalla (ed.). Technology and Employment
in Industry: A Case Study Approach. Geneva: International Labour
Office. pp. 255-274.
Brandt, Hartmut; Elbert Gerken and Bernd Schubert. 1972. The Industrial
Town as a Factor of Economic and Social Development: The Example
of Janja, Uganda. Munchen: Weltforum, Verlag.
Broehl, Wayne Jr. 1978. The Village Entrepreneur: Change Agents in
India's Rural Development. Cambridge, Massachusetts: Harvard
Byerlee, Derek; Carl Eicher, Carl Liedholm and Dunstan Spencer. 1977.
"Rural Employment in Tropical Africa." African Rural Employment
Working Paper No. 20. East Lansing: Michigan State University,
Department of Agricultural Economics.
1979. "Employment-Output Conflicts, Factor Price Distortions
and Choice of Technique: Empirical Results from Sierra Leone."
African Rural Economy Working Paper No. 26. East Lansing: Michigan
Byerlee, Derek, Joseph L. Tommy and Habib Fatoo. 1976. "Rural Urban
Migration in Sierra Leone: Determinants and Policy Implications."
African Rural Economy Paper No. 13, East Lansing: Michigan State
University, Department of Agricultural Economics.
Carroll, J. J. 1965. The Filipino Manufacturing Entrepreneur: Agent
and Product of Change. Ithaca: Cornell University Press.
Cartillier, Michel. 1975. "Role of Small-scale Industries in Economic
Development." Economic and Political Weekly. (November 1).
Chenery, Hollis, Montek S. Ahluwalia, C. L. G. Bell, John H. Duloy and
Richard Jolly. 1974. Redistribution with Growth. London: Oxford
Child, Frank C. 1973. "An Empirical Study of Small-Scale Rural Industry
in Kenya." IDS Working Paper No. 127. Institute of Development
Studies. University of Nairobi, Kenya.
Child, Frank C. and Hiromitsu Kaneda. 1975. "Links to the Green Revolution:
A Study of Small-scale, Agriculturally Related Industry in the
Pakistan Punjab." Economic Development and Cultural Change 23:249-275.
Chuta, Enyinna. 1978. "The Economics of the Gara (Tie-Dye) Cloth
Industry in Sierra Leone." African Rural Employment Working Paper
No. 25, East Lansing: Michigan State University, Department of
1979. "The Bread Industry of Sierra Leone: A Case Study in
Appropriate Technology." East Lansing: Michigan State University,
Department of Agricultural Economics. (Mimeographed).
Cochran, T. C. 1959. The Puerto Rican Businessman. Philadelphia:
University of Pennsylvania Press.
Davies, Omar; Yacob Fisseha, Annette Francis, and C. Kirton. 1979.
"A Preliminary Analysis of the Small-Scale, Non-farm Sector in
Jamaica." Small-Scale Enterprise Survey Unit. Institute of
Social and Economic Research, University of the West Indies,
De Vries, Jan. 1974. The Dutch Economy in the Golden Age, 1500-1700.
New Haven: Yale University Press.
Dhamija, Jasleen. 1976. "Non-Farm Activities in Rural Areas and Towns:
The Lessons and Experiences of Iran." Studies in Employment and
Rural Development No. 31. Washington, D. C.: World Bank.
Dhar, P. N. 1958. Small Scale Industries in Delhi. Bombay: Asia
Dhar, P. N. and H. F. Lydall. 1961. The Role of Small Enterprises in
Indian Economic Development. Bombay: Asia Publishing House.
Eckhaus, R. S. 1955. "The Factor Proportions Problem in Underdeveloped
Areas." American Economic Review 45:539-565.
El-Karanshawy, H. A. S. 1975. "Choice of Carpet Weaving Technology in
Egypt." Ph.D. dissertation, University of Strathelyde.
Falcon, Walter P. 1967. "Agricultural and Industrial Interrelationships
in West Pakistan." Journal of Farm Economics 49:1139-1157.
Gadgil, D. R. 1959. Origins of the Modern Indian Business Class An
Interim Report. New York: Institute of Pacific Relations.
Galenson and Leibenstein. 1955. "Investment Criteria, Productivity and
Economic Development." Quarterly Journal of Economics, 69:343-370.
Gaude, J. 1975. "Capital-Labour Substitution Possibilities: A Review
of the Empirical Evidence," in Technology and Employment in
Industry, (ed.) A. S. Bhalla. Geneva: I.L.O.
Georgia Institute of Technology. 1975. "An International Compilation
of Small Scale Industry Definitions." Atlanta, Georgia: Georgia
Institute of Technology, Engineering Experiment Station (Mimeographed).
Gerken, E. 1973. "Land Productivity and the Employment Problem of Rural
Areas," Center Discussion Paper No. 176, New Haven: Yale University.
Gibb, Arthur. 1974. "Agricultural Modernization, Non-Farm Employment
and Low Level Urbanization: A Case Study of a Central Luzon Sub-
Region of the Philippines," Ph.D. dissertation, Ann Arbor: University
Hagen, Everett. 1962. On the Theory of Social Change: How Economic
Growth Begins. Homewood, Illinois: Irwin.
Haggblade, S.; Jacques Defay, and Bob Pitman. 1979. "Small Manufacturing
and Repair Enterprises in Haiti: Survey Results." Michigan State
University Rural Development Series, Working Paper No. 4, Department
of Agricultural Economics, East Lansing: Michigan State University.
Harper, Malcolm. 1978. Consultancy for Small Business. London: Inter-
mediate Technology Publications. Ltd.
Harris, John R. 1970. "Nigerian Entrepreneurship in Industry," in Growth
and Development of the Nigerian Economy, edited by Carl Eicher and
Carl Liedholm. East Lansing: Michigan State University.
Hawbaker, George and Howard H. Turner. 1972. "Developing Small Industries:
A Case Study of A.I.D. Assistance in Nigeria, 1962-1971." Washington,
Hirschman, A. E. 1958. The Strategy of Economic Development. New Haven:
Yale University Press.
Ho, Samuel P. S. 1976. "The Rural Non-Farm Sector in Taiwan," Studies
in Employment and Rural Development No. 32. Washington, D.C.:
1978. Economic Development of Taiwan, 1860-1970. New Haven:
Yale University Press.
Hoselitz, Bert F. (ed.). 1968. The Role of Small Industry in the Process
of Economic Growth. The Hague: Mouton.
Houthackker, H. S. 1961. "An International Comparison of Personal Savings."
Bulletin of the International Statistical Institute, 38:55-69.
Huddle, Donald. 1977. "An Analysis of the Savings Behavior of a Group
of Colombian Artisan Enterpreneurs," Program of Development Studies
Paper No. 81. Houston, Texas: Rice University.
Huddle, Donald L. and Thi-Min Ho. 1972. "The Contribution of Traditional
and Small-Scale Culture Goods in International Trade and Employment,"
Program of Development Studies Paper No. 35. Houston, Texas: Rice
Hymer, Stephen and Stephen Resnick. 1969. "A Model of an Agrarian Economy
with Nonagricultural Activities," The American Economic Review,
India, Government of. 1965. National Sample Survey No. 94. Calcutta,
India: Cabinet Secretariate.