EMPLOYMENT AND GROWTH IN SMALL-SCALE INDUSTRY
Empirical Evidence and Policy Assessment from Sierra Leone
Enyinna Chuta and Carl Liedholm
St. Martin's Press New York @ International Labour Organisation 1985 To ourparents All rights ieserved. For information, write: St. Martin's Press, Inc., 175 Fifth Avenue, New York, NY 10010 Prin'ed in Hong Kong Published in the United Kingdom by The Macmillan Press First published Ltd. in the United States of America in 1985 ISBN 0-312-24458-4
Library of Congress Cataloging in Publication Data Chuta, Enyinna. Employmezt and growth in smal!-scale industry. Includes index. I. Small business--Sierra Leone. 2. Labor supply- Sierra Leone. I. Liedholm, Carl. II. Title. HD2346.S5C46 1985 338.6'42'09664 84-24771 ISBN G- 3 12-24458-4 Contents
xi List of Tables xv List of Figures xv Foreword Preface 1 I INTRODUCTION Small-scale Industry PART I The Role of Rural and Urban in Employment Generation
PROFILE OF SMALL-SCALE 2 DESCRIPTIVE 13 INDUSTRY IN SIERRA LEONE Sector 13 2.1 Overall Importance of the Small-scale Industry Relative Importance of Categories of Small-scale 2.2 17 Industry 22 2.3 Annual Output and Value-added 23 2.3.1 Seasonal Variations in Output 27 2.4 Material and Services Input and Stock Inputs 28 2.5 Capital Services 31 2.6 Initial Capital Stock Requirements 35 2.6.1 Sources of Initial Capital 37 2.6.2 Sources of Expansion Capital Excess Capacity 37 2.6.3 40 2.7 Labour Input 45 2.7.1 Characteristics of Apprentices 48 2.7.2 Characteristics of Proprietors
3 DETERMINANTS OF DEMAND FOR AND SUPPLY SMALL-SCALE INDUSTRY OUTPUT AND OF 52 EMPLOYMENT IN SIERRA LEONE 52 Demand 3.1 57 3.2 Supply Contents viii Contents Microeconomic Analysis of Gara Firnib 103 Production Processes and their 6.3 3.2.1 Alternative 6.3.1 Efficiency of Reptesentative Types./ 103 Factor Proportions 57 6.4 Marketing of Gara Cloth 109 3.2.2 Variations in Factor Proportions by Location 62 109 64 6.4.1 Domestic Marketing 3.2.3 Variations in Factor Proportions by Firm Size 110 Profit 6.4.2 International Marketing 3.2.4 Returns to Proprietor and 'Economic' Marketing 110 65 6.4.3 The Co-operatives and Gara Rate 111 Entrepreneur 70 6.5 Policy Implications 3.5 The Role of the 6.5.1 Production Polic;es 111 6.5.2 Marketing Policies 111 4 GROWTH ASPECTS OF SMALL-SCALE INDUSTRY 75 113 4.1 Aggregate Enterprise and Employment Growth (1974-80) 75 7 THE BREAD-MAKING INDUSTRY Employment Growth Patterns (1974-80) 79 4.2 Sectoral 7.1 Baking Technology in Sierra Leone 113 Closure Rates 80 4.3 Entry and 7.1.1 Major Types of Baking Equipment 113 of Innovation 82 4.4 Patterns 7.1.2 The Distribution of Oven Types and Rein,'estment 83 115 4.5 Patterns of Equipment 116 of Bread-making in Sierra Leone 7.1.3 The Process 120 PART II Some Empirical Case-studies 7.2 The Economics of Bread-making 123 7.2.1 Microeconomic Analysis of Bakery Firms 5 THE CLOTHING INDUSTRY 87 7.2.2 Factor Proportions among Bakery 124 88 Representative Firm Types ' 5.1 An Overview of the Sierra Leone Clothing Industry Technical Efficiency among Bakery Firms 89 7.2.3 5.2 Budget Analysis of Representative Firm Types 124, among Representative Firms 91 Representative 5.2.1 Technical Efficiency profitability of Representative Bread among Representative Firms 92 7.2.4 Economic 5.2.2 Economic Efficiency making Firms 124 5.2.3 Economic Profits among Clothing Firms 92 Pattern of Technology Adoption in the Sierra Leone and Implications of Analytical Results 95 7.3 5.3 Summary Bread Industry 126 Quality 127 96 7.4 Lccation Factors, Bread Technology and 6 THE GARA (TIE-DYE) INDUSTRY 128 7.5 Comparative Static Analysis of Sierra Leone Bakeries of Gara Dyeing 96 130 6.1 The Technology 7.6 Siczra Leone Bakeries in the African Context 132 6.1.1 Evolution of Gara Dyeing in Sierra Leone 97 7.7 Implications of the Bakery Case Study 6.1.2 The Use of Synthetic Dyestuff 97 6.1.3 Methods Used in Producing Patterns 97 PART III Policy Assessment 6.1.4 Designs 99 The Dyeing Process 99 A REVIEW OF POLICY ENVIRONMENT, 6.1.5 137 6.1.6 Finishing 99 PROGRAMME AND INSTITUTION SUPPORT and Quantity Control Problems 100P137 6.1.7 Quality Economic Policies 137 Inputs 101 8.1 6.2 Production 8.1.1 Credit Policies 137 6.2.1 Labour 101 102 8.1.2 Policies for Delivering Technical and 6.2.2 Capital Managerial Assistance 139 Material Inputs 102 6.2.3 8.1.3 Fiscal Policy 140 6.2.4 Management Inputs 103 x Contents
8.1.4 Infrascructure Policies 142 8.1.5 Large-scale Industrial Policies 143 8.1.6 Agricultural Policies 144 8.2 -'A Review of Existing Programme and Institutional S u p p rt 145 • 8.2.1 Current Institutional and Programme List of Tables Arrangement 146
9 SUMMARY AND MAJOR POLICY IMPLICATIONS 2.1 Sierra Leone: Distribution of industrial establishments OF THE STUDY 148 by location and size, 1974 14 9.1 Summary of Major 2.2 Sierra Leone: Estimates of value-added by large- and Findings 148 small-scale industry 9.2 Major Policy Conclusions 16 151 2.3 Sierra Leone: Distribution of small-scale establishments Not and References 153 and employment by industrial catcgories and size of Selective Bibliography 169 location, 1974 19 Index 1"/5 2.4 Sierra Leone: Value-added by small-scale industrial categories by location, 1974-5 21 2.5 Sierra Leone: Annual mean output and value-added by major small-scale industrial categories and size of locali ties, 1974-5 24 2.6 Sierra Leone: Annual mean capital services input by major small-scale industrial categories and size of locali ties, 1974-5 32 2.7 Annual mean capital stock for five major small-scale industrial categories by size of locality, 1974-5 34 2.8 Sierra Leone: Average initial capital stock requirements at time of establishments by major industrial categories and size of locality, 1974-5 35 2.9 Sierra Leone: Sources of initial capital of small-scale industrial firms at the time of establishment 36 2.10 Sierra Leone: Sources of expansion funds for small-scale industrial establishments, 1974-5 38 2.11 Sierra Leone: Excess capacity by major small-scale in dustry by location, 1974-5 39 2.12 Sierra Leone: Annual mean labour hours by small-scale industrial category, labour type and locality 42 2.13 Sierra Leone: Duration of apprenticeship and learning fee by small-scale industrial category and location 46 2.14 Sierra Leone: Percentage of small-scale industry pro prietors charging a learning fee to apprentices 47 xi xii List of Tables List of Tables Xi 2.15 Sierra Leone: Percentage of small-scale industrial pro- 7.2 The distribution of bakeries prietorg' fathers who were farmers by industrial category by type of equipment and and location size of locality 4 117 7.3 Processes of bread-making in Sierra:-Leone, 2.16 Sierra Leone: Percentage of small-scale industrial 1974-5 119 pro- 7.4 Fixed costs among prietors with formal education by industry representative bakery firns in Sierra and location 49 Leone, 1974-5 2.17 Sierra Leone: Percentage of small-scale 122 industrial pro- 7.5 Factor proportions prietors who keep records among represenlative bakery firms in 50 Sierra Leone, 2.18 Sierra Leone: Age of small-scale 1974-5 123 industrial establish- 7.6 Budgets ments, 1974 of representative bakery firms in Sierra Leone, 51 3.1 1974-5 Total expenditure elasticities of demand by rural 125 house- 7.7 Growth rates in the holds for major small industry products, number of bakery establishments 1974-5 54 and employment 3.2 Sierra Leone: Mean values of the output-capital, in some localities of Sierra Leone, 1974 output- and 1980 labour and labour-capital ratios by process for 128 major 7.8 Observed energy costs in baking in five bread-making small-scale industry categories, 1974-5 60 enterprises 3.3 Sierra Leone: - Sierra Leone, 1980 Output-capitai, output-labour, 129 capital labour ratios for traditional small-scale industries by industry and location, 1974-5 63 3.4 Sierra Leone: Annual mean return to proprietor and economic profit rate by major process and major small scale industry category 3.5 67 Sierra Leone: Annual returns to proprietor by major industry and locality for traditional processes only (Leones per year) 68 4.1 Employment by branch of activity as percentage of total small-scale manufacturing, 1980, and compound annual growth rates in employment and number of establish ments in small-scale manufacturing by size of locality, 1974-80 76 4.2 Percentages of firms of various sizes by size of locality, 1974 and 1980 79 4.3 Innovations introduced by small-scale enterprises of at least five years' standing in Sierra Leone, 1975-80 83 5.1 Factor proportions for the clothing representative firm types 90 5.2 Budgets of the representative clothing firms 93 6.1 Representative firm types of gara firms in Sierra Leone, 1974-5 104 6.2 Factor proportions for the gara representative firm types 105 6.3 Budgets of gara dyeing representative fir-1 types, 1974-5 107 7.1 Sample ar d population distribution of bakery KI by size of oven types locality - Sierra Leone, 1974-5 116 List of Figures Foreword
2.1 Sierra Leorne: Index of monthly output of major small- Agricultural sectors in scale industries, 1974-5 developing countries cannot absorb the rap 26 idly growing rural labour force. Neither can their modem urban sectors contribute significantly to employment generation. It is this t recognition that has brought o the fore the need to promote rural non-farm employment opportunities within the framework of na tional development strategies. In spite of the growing plethora of literature on rural industrialisation, however, not much is known about the employment and income potential of non-farm activities. The present book by Dr Enyinna Chuta and Profossor Carl Lied holm on small-scale industry in Sierra Leone is one of the few examples of a systematic analysis of growth potential of rural small scale activities. The analysis is undertaken for two points in time, 1974-5 and 1980, thus enabling an examination of income and em-, ployment growth over time. The study concludes that the small-sale activities are not only labour-intensive but they also generate respect able rates of profits. It further suggests that the scope for employ ment promotion through rural small industries cannot be tapped effectively unless relevant policies are suitably modified to account potential take into conflicts between large-scale and small-scale and urban and rural industries. The study was undertaken as part of research on rural industrialisa tion and related non-farm activities within the framework World of the ILO Employment Programme. In the past the ILO has sponsored undertaken cross-country studies in the field of rural industrialisa tion, and appropriate technologies. It is also engaged in providing technical advisory services to member governments implementing on request, and technical co-operation projects in the promotion of rural and urban small-scale industries. An individual country case study such as this one on Sierra Leone adds depth to the cross-country existing knowledge base, thus highlighting methodological and policy issues that may be relevant to other developing countries interested in promoting rural small industry as an instrument of employment policy. xv xvi Foreword The bulk of the field work for this case-study was financed by the United States Agency for International Development (USAID) through Michigan State University, with which Professor Carl holm is associated. Lied The second field survey and writing up chapters of several of the book were undertaken by Dr behalf Enyinna Chuta, on of the ILO when he was a staff member of the Technology and Preface Employment Branch. The ILO would like to Michigan State University express its gratitute to the USAID and for their collaboration in the study. This book reports on the findings of a comprehensive study of small-scale A. S. BHALLA industry in Sierra Leone. Three be highlighted at the outset. features of the book may First, it provides the first detailed picture Chief of the entire Technology and Employment spectrum of small-scale industries in Sierra Branch including those located Leone, in rural areas. In view of the wide recognition1 of the inadequate knowledge base to buttress small-scale programmes in most developing industry countries, this book helps fill an important information gap. Second, the book describes the rather unique survey methodology that was employed to generate the data from the small enterprises, the majority of which kept no books or records. 2 That methodology included the use of repeated (twice weekly) interviews of a panel of entrepreneurs over a f, ll-year period. Third, by also undertaking a review of Sierra Leone policles ahd programmes for promoting small industry and handicrafts, the volume seeks to highlight the important link between empirical information and its use in the design and implementation of policies, programmes and projects. It is envisaged that this book may be of value planners, financial to policy-makers, institutions, researchers, aid donors, international agencies, voluntary organisations and local institutions involved promoting small-scale enterprises in and handicrafts. It may also prove to be a useful complement to existing studies of small-scale industry in Nigeria and Ghana in the 3 West African region and may also be of assistance to those planning similar studies in other Third World countries where information on small-scale industry is deficient. The research on which the Sierra Leone financed study is based was mostly by the US Agency for International Development with Michigan contracts State University (AID/csd 3625 and AID/ta-CA-2, and DAN-1090-A-00-20 8 7 -0). The assistance of Dr Cliff Barton and Dr Michael Farbman is particularly deserving of mention. The ous sulrveys were undertaken vari in collaboration with Njala University College of the University of Sierra Leone. Both authors were resi dent in Sierra Leone during the major survey period of 1974 to 1975 xvi xviii Preface
as well as during the summer of 1980 when a follow-up survey was undertaken. This second survey was jointly financed by the ILO and the Development Support Bureau, AID, Washington, DC. We are grateful to the Government of Sierra Leone for approving and patronising the small-scale industry surveys. The unflinching support received from Mr. M.A. Tunis, Head of the Small-Scale Introduction Industry Division, Ministry of Trade and Industry, was invaluable. We are also grateful to Dr. S. Matura. and Professor J.Kamara, former and present principals of Njala Jniversity College during the Rapid industrialisation has for several decades been regarded as an 1974-5 and 1980 surveys respectively, for approving the research attractive and effective strategy for transformirg developing collaboration between Njala University College and Michigan econ State omies and maximising their rates of economic growth. Indeed, many University. The direction and support offered by Dr. Dunstan Spen- developing countries began in the early 1950s to launch cer, leader of the Michigan State University research industrialisa team at Njala in tion drives based generally on an import-substitution strategy, 1974-5 are gratefully acknowledged. Needless which to mention, we appre- usually resulted in the establishment of large-scale capital-intensive ciated all the useful suggestions and assistance offered by other team industries in urban areas. In recent years, however, it has become members sucii as Professor Derek Byerlee (now of the CIMMYT apparent that this strategy has often produced disappointing results. Mexico), Messrs. Ibi May-Parker (now at the Bark of Sierra Leone), In a number of developing countries not only has the over-all rate James Kamara and Ola Robert. The surveys would of have been imposs- growth of the economy been low but cmployment in the ible without the hard work and industrial devotion of our entire field staff sector has failed to keep pace with population (enumerators, supervisors growth and in some and drivers), the co-operation of respon- cases even declined in absolute terms.' dents and the support of local The poor employment per of them. chiefs and elders. We are grateful to all formance has been the cause of particular f r ac a e n te cu e o atc lr concerncn e nb because c u e employe p o, ment generation has recently become an independent policy goal In the final revision of manuscript we have benefited from the many countries.2 for incisive comments of Ajit Bhalla, S.V. Sethuraman, Susumu Wata- Many developing countries have consequently become increasingly nabe, Frank Lisk, Ghazi M. Faroog and Lowell Taylor. aware of and interested in assessing the role that small-scale establish
ENYINNA ments might play in their industrialisation strategies. If, for example, CHUTA small-scale industrial establishments are more labour intensive, more CARL LIEDHOLM widely dispersed, generate more output per unit of capital, require less foreign exchange, and produce a higher 'economic' profit than their large-scale counterparts, then strong economic justification would exist for promoting small-scale firms. Sierra Leone's experience with industrialisation has been similar to that of other developing countries. After independence in 1961, Sierra Leone followed an import-substitution strategy in which poli cies were designed to encourage the expansion of large-scale urban based firms. The results of this policy proved to be disappointing. The manufacturing sector grew at a real annual rate of only 2.8 per cent from 1965-6 to 1971-2, while during the same period the number of individuals employed in 'large-scale' manufacturing firms actually declined at a compound annual rate of 3.5 per cent.3 More over, by 1971 unemployment in urban areas had grown to almost 14
1 4 Introduction Introduction 5
in order to address the various issues already discussed, a combina- made of every industrial establishment operating in these twenty-four tion of different data collection strategies were utilised for obtaining enumeration areas. the required information. Some details of our survey methodology In each of the localities and 'enumeration areas' examined during will now be presented. phase I, the enumerator- obtained the following information about each establishment:
1.3 SURVEY METHODOLOGY (a) the type of activity; (b) the number of workers, including the proprietor; hired work It is important to state from the outset that two separate surveys are ers and apprentices: involved in the Sierra Leone study, the first in 1974-5 and a follow-up (c) the type of workshop (whether temporary, mud or cement); survey which took place in 1980. (d) the number and kind of machines used (whether power-driven or manually powered).
1.3.1 1974-5 Survey These data were rather quickly obtained during the street-by-street enumeration undertaken in each of the selected localities and 'enu The 1974-5 survey was undertaken in two phases. During phase I of meration areas'. Fortunately, excellent cooperation was obtained the data collection, a census of Sierra Leone's small-scale industrial from the local authorities in these areas, and in most cases they sector was undertaken to provide an estimate of the population of provided the enumerators with local personnel who assisted in locat small-scale industrial establishments in the entire country. Such an ing the various establishments. Indeed, since most of the industrial endeavour was unique because it represented one of the first at- establishments did not have signs and were often hidden behind other tempts in Africa to estimate t, total number of industrial enterprises buildings, these local informants were vital in ensuring that 'an in both the rural and urban areas of one country." In view of accurate enumeration of firms was obtained. budgetary and time limitations, however, it was not possible to Finally, it should be noted that the phase I data were collected undertake a complete enumeration of every industrial establishment during the period from March to June 1974 - a period during which in Sierra Leone. Thus, it was necessary to devise a stratified sampling one might have expected a sizeable variation in the level of non-farm procedure to obtain an estimate of the establishment population. aciivity. This variation would have been expected because the 'wet' Since previous empirical studies have revealed that both magnitude season in Sierra Leone normally begins in early May, at which time and composition of industrial activity varies with size of settle- non-farm activity would be hypothesised to decline as a result of the ments," the stratification was based on locality size. marked increase in the demand for farm labour for planting. Fortu The following procedure was thus used to obtain an estimate of the nately, for the present study, the 'wet' season in 1974 did not begin total number of industrial establishments. First, a total establishment until the end of June, and thus the results of the phase I survey should enumeration was undertaken in the eighteen localities, including provide a good estimate of the population of rural small-scale indus Freetown, with populations in excess of 20 000.12 Secondly, one half trial establishments during a period when they would be expected to of the twenty-four localities with 2000-20 000 inhabitants were ran- be reasonably active. domly selected for total establishment enumeration.3 Finally, itwas The phz.se II survey was designed to generate more detailed necessary to obtain an estimate of the number of industrial establish- economic data from a selected sample of industrial establishments ments in the smallest census units, the 'enumeration areas'.' 4 To over a one-year period - August 1974 to July 1975. The results of the ensure consistency with the other micro-level components of the phase I survey were used to assemble the frame from which the Sierra Leone Rural Employment Research Project, twenty-four rural sample of firms could be chosen. Two-thirds of the 270 firms selected 'enumeration areas' selected for these studies were used for the in those localities with 2000 or more inhabitants were chosen on a small-scale industrial component as well. Thus, a total count was random basis to ensure that a reasonable approximation of the 6 Introduction Introduction 7 undelying population would be obtained.' 5 The remaining one-third The entire survey lasted four weeks. In each of the locdities and were chosen in a purposive fashion to ensure that a complete spec- enumeration area villages, the enumerators obtained information on trum of production techniques could be obtained, the type of industrial activity, the number of yorkers (including Two major types of data were collected during the phase II survey. proprietors, hired apprentices and children), the type of workshop Firstly, stock data were obtained at the beginning and end of the (temporary, permanent, etc.), the number and kind of machines used survey period for buildings, tools, equipment and furniture, material (whether power-driven or manually operated).: All establishments inputs and outputs. Secondly, data on sales, output, labour used and were listed in each locality but only every fifth enterprise was system material inputs purchased were collected twice weekly throughout a atically selected for administering a second questionnaire designed one-year period. to ascertain what changes had taken place with respect to each enter Finally, we must point out that the 1974-5 survey of small-scale prise's workshop layout, method of producing goods and services and industries was undertaken as part of an integrated Rural Employ- product types, banking and bookkeeping practices, and profit rein ment Research Project based at Njala University College, Sierra vestments between 1975-6 and 1979-80. Both questionnaires were Leone. Other components of the Sierra Leone research project are pre-coded and pre-tested before being utilised for the actual survey. farm-level, consumption, marketing, migraion and food-processing Finally, on the basis of the sample frame that was assembled in studies. The different study components were co-ordinated under 1974, an attempt was made to rediscover in 1980 some firms which one leadership, utilising almost the same rural households, field were studied during the 1974-5 phase II survey in each of the survey workers and vehicles and were completed within the same time localities. The idea was to ascertain the rates of growth and attrition frame. It should not therefore be surprising when we make refer- among small enterprises in Sierra Leone. In some cases where these ences to survey data other than the small-scale industry data, the firms were not located, an attempt was made to ascertain if the firm collection of which was our primary responsibility, had closed down, moved to a different part of the city or to an entirely different locality.
1.3.2 1980 Survey 1.3.3 Sources of data for the case studies Following the procedures of the 1974-5 phase I street-by-street enumeration of small-scale industries in Sierra Leone, a stratified The input-output data used in the case study chapters were generated sampling procedure based on locality size was adopted during a by the authors during the 1974-5 survey of small-scale industries in follow-up survey in 1980. Thus, a total establishment enumeration Sierra Leone. Different subsamples (random and purposive) of 132 was undertaken in the central constituency of Freetown, and it was tailoring firms, 22 gara (tie-dyed cioth) firms, and 31 bakery firms, used as a sample for the Greater Freetown area, which has a popula- were utilised for the three case studies reported in Part II of this tion of over 250 000. Three other cities, namely Bo, Makeni and volume. For the purpose of the case studies, questionnaires were Kenema, with a population of between 20 000 and 250 000 were designed to obtain additional in-depth information on the existing totally enumerated. A total establishment count was also undertaken technology of production, organisation, and marketing aspects of in Mattru, Port Loko and Segbwema, which fall into the third stratum each industry. The information on the two large-scale clothing firms of localities with a population range of 2000-20 000. Finally, it was was based on secondary data.16 necessary to obtain some estimate of the number of industrial estab lishments in the smallest census units - the enumeration area villages. Such villages include Bomabali village, Mateti, Tikonko, Taninahun, 1.4 AN OUTLINE OF THE BOOK Kpandebu and Lalehu. Those 1980 survey localities were also enu merated in 1974 and they were distributed throughout the four This book has been written in three parts. In Part I, we present both administrative areas of Sierra Leone. descriptive and analytical information relating to the role of rural and 8 Introduction Introduction 9 urban small-scale industries in employment generation. Part II con- making industry case study tains detailed case study highlights theneed to upgrade traditional analysis of alternative techniques of produc- policytion in environmentthree major small-scale industries. baking technology. ho and pormesprtareIn Part III, assessments of Chapters 8 and 9 of Part III are fol-l Chptrn8anerotarakearn, respectively devoted to a review of lowed by a summary of programme support ariundertaken, fol- dvte o rvewo ofthestudy and a note of its policy implications, policy environment, programme and institutional Chapter 2 of Part I provides a descriptive summary of the study and a note of its inaplications. support and a profile of small industries review in some In Chapter 8, we in Sierra Leone, covering information on utilisation detail Sierra Leone's small industry promotion efforts of labour and as outlined in the 1974 capital services, output and value-added generated NationalDevelopment Planand some aspects by small-scale of the existing institutional industries by subsectors and location (rural and framework. Finally, in Chapter 9 we urban), as well as highlight some of some entrepreneurial characteristics. In Chapter the policy implications of the study. At the policy 3,the major deter- level, we indicate minants of demand for and supply of small-scale that there is a need for separate legislation which examined. industry output are clearly specifies measures On the demand side, we examine three major for promoting small enterprises and handi demand sources of crafts. Such measures should for small-scale industry products. Thus information embrace not only direct policies (such relating as credit, fiscal, etc.) but to expenditure elasticities, forward and backward also indirect measures such as the introduc linkages and for- tion of more favourable eign demand for small enterprise goods are presented. income-generating agricultural policies. On the supply the level of individual At side, alternative production processes and firms, programmes are needed to upgrade their factor proportion traditional technologies, ratios are examined as well as the returns improve skill levels and therefore productiv to the proprietor and the ity of resources, economic profits earned by the various update managerial practices, transform the tradi small-scale industry groups. tional apprenticeship On the basis of re, ssion analysis system into a useful vehicle for transferring we highlight factors which deter- modern mine entrepreneuriii success in Sierra technical and managerial skills, and eliminate Leone small-scale industry. A constraints marketing comparative static analysis of employment (at the domestic and international level). growth, patterns and existing Given the innovation and reirvestment of profits information base in Sierra Leone, and the overall over the period 1974 to 1980 development national has been undertaken in Chapter 4 to highlight objectives, it may not be too difficult to design some issues of viability priate policies appro in Sierra Leone small-scale industry. Of particular and programmes for developing small-scale enter interest is the prises and handicrafts. finding of differential rates of employment In this respect, the study shows that growth between different important it is industrial subsectors and localities to tailor intervention measures to the particular of different population sizes. Inno- circumstances needs and vative behaviour of small producers and products. also varied between industrial subsectors but rate of reinvestments the of past profits in current business was found to be disappointingly low. In Part 11, Chapters 5 to 7 have been respectively devoted to case studies of three of the five major small-scale industries, namely, the clothing, gara (tie-dye) and bread-making industries. try In each indus category, various technologies being utilised during period the survey have been identified. The analysis of factor economic proportions, and technical efficiency and economic profitability undertaken are in order to compare the viability of the logical various techno types. The policy implications of each case study quite analysis are unique. For example, in the gara industry, improvement production in methods is crucial; whereas for small tailors, stabilisation of the machine rental market and provision of the necessary for the purchase capital of sewing machines is quite important. The bread- Part I The Role of Rural and Urban Small-scale Industries in Employment Generation 2 Descriptive Profile of Small-scale Industry in Sierra Leone
This chapter will provide a profile of the extent, nature and growth of small-scale industry in Sierra Leone. Such a descriptive overview can usefully serve to provide a foundation for understanding the role played by small-scale industries in that country. Most of the back ground information presented in this chapter relates to 1974, the year our detailed, industrial survey was undertaken. The overall magni tude of the small-scale industry sectcr will be examined first, followed by a discussion of the relative and absolute importance of the indi vidual industries. Seasonal variations in the output levels will'be treated next. The material. capital and labour inputs of the various industries will then be examined in some detail.
2.1 OVERALL IMPORTANCE OF THE SMALL-SCALE INDUSTRY SECTOR
One of the most important findings of the research is the discovery that the industrial sector' in Sierra Leone is much more extensive than had been previousiy recognised. This is apparent from Table 2.1 which presents the distribution of industrial establishments and in dustrial employment by location and size. These figures reveal that in 1974, approximately 50 000 industrial establishments employed al most 93 000 individuals. The Government of Sierra Leone on the other hand had assumed that only 52 000 individuals would be employed in this sector in 1974.2 Thus, the phase I survey results indicate that the Government's industrial employment estimate should have been increased by three-quarters. The surveys also reveal that small-scale industrial establishments dominate the industrial sector in terms of both the number of firms 14 Small-scale Inaustry in .3erraLeone 15
00 t r3' r- I E and total employment. 'Small scale' was defined in this study to o6 _ include those establishments employing fewer than 50 persons.3 -" = .i Using this definition, there were only twenty ei ht large-scale estab c E-= lishments in Sierra Leone, and those firms mployed only 4111 4 0N *-. individuals. In those localities with populations in excess of 2000 on M, ,-' the other hand, there were approximately 5000 small-scale industries - k 0. employing approximately 15 000 individuals (see Table 1.1). More " .-E . ""over, it was projected that there were approximately 45 000 small scale establishments employing 73 000 persons in those localities with 5 -E populations less than 2000. Thus, these results indicate that small C; ,;,- , -C scale establishments account for over 95 per cent of the employment 4?- Ent.,- in Sierra Leone's industrial sector. E0M>. The average size of the industrial establishment in Sierra Leone S- was very small. The 'average' industrial firm, for example, employed r , 00 n en 2only 1.8 workers.' Indeed, 98.0 per cent of the firms employed less .. CDen - -C than 5 individuals, 0.7 per cent employed from 5 to 9 individuals, 0.3 - ~ 0 - 0 E "per cent employed from 10 to 49 individuals and only 0.1 per cent employed over 50 individuals. 7 Thus, in terms of employment, the vast majority of firms in Sierra Leone were concentrated at the t = - .o lowest end of the small-scale industry size continuum. - ,Another useful indicator of the relative importance of small-scale - E industrv is 'value added'. In our 1974 Sierra Leone study, the annual 0.0 - -. value added of the small-scale industrial establishments has been estimated from the output and cost dat, collected from firms on a -a = .twice-weekly enumeration. The mean value added Jor each of those >~ ., .firms in the smail-scale industrial category in each locality group was L-CJ (U- calculated and then multiplied according to the number of establish _ ments in each area.' The resulting estimates for small-sr-ale industry - 6 U UV as well as the various estimates for small- and large-scale industry E E prepared by the Government of Sierra Leone are summarised in _i _ -, -0 Table 2.2. " 0 .0 :F0Table 2.2 reveals that small industry in 1974-5 accounted for .2. .0_R7 E,-! r" M C approximately Le. 13 million9 2.9 wu . or per cent of Sierra Leone's Gross . EE .. Domestic Product. This estimate is surprisingly close to the value -< "- .....--. -f- -..0. __ 0 0. .. added estimate prepared by the Central Statistics = ". National Accounts, an estimate which was not based onOffice an extensive for the iz Q " r E5o.- -' '--- .When survey andthe which value wasadded admittedly for small-scale 'very rough'. industry is compared with o UV Lo ,,, the estimate of the value added of large-scale industry (Le.17.4 0 0001J- 0 0 0 t3 E 0 1 -1.1--- million r ,, or 3.9 per cent of Sierra Leone's Gross Domestic Product in 14 . -n- 1974-5), the small-scale establishments are revealed to contribute 16 Small-scale Industry in Sierra Leone 17
E , approximately 43 per cent of the entire industrial sector's value 6 a added. Although the value added percentageqof small-scale (43 per CZ> > cent) is much lower than the percentage empldyed in small-scale (i.e. " 9 95 per cent) relative to large-sczale industry, the study shows that E . small-scale establishments are a significant component of Sierra , -. ,.Leone's industrial sector.
- 4. r-q, U - Moreover, the small-scale component of Sierra Leone's indus "-, " fn trial sector appears to be relatively large when compared with other developing countries. Data on the very small industrial firms in most countries are not generally collected and thus, of necessity, only a limited number of comparisons are possible. Small-scale industry CC employing less than 50 workers accounted for 95 per cent of the manufacturing employment in Sierra Leone, but only 70 per cent in -,i i t; .Nigeria," ° 71 per cent in Tunisia," 79 per cent in the Philippines 2 and 3 =. . 69 per cent in Colombia. The corresponding contribution of small scale industries to value added in manufacturing, on the other hand, was 43 per cent in Sierra Leone, 32 per cent in Colombia and only 14 C.) ,U-- per cent in the Philippines. In these countries, the percentage contri-
• . bution of small-scale industries to value added was thus substantially S. _ less than employment; indeed, the value added contribution was less than one-half that of employment. '- In terms of both employment and value added, however, the 4 = o relative importance of small-scale industry in Sierra Leone appears to Q-- be somewhat greater than elsewhere. This result is perhaps traceable in part to the extensive effort made in our Sierra Leone survey to include the small rural establishments, which are often overlooked, 0. 0= and thus are underestimated in surveys. There may also be economic -- -o reasons, such as the small size of the Sierra Leone domestic market, - .that account for the relatively large numbers of very small-scale . .industrial enterprises in Sierra Leone. Finally, some of the official .o 0 ~figures for other countries may not include part-time workers or ,,.Z E o apprentices, and so the data may not be perfectly comparable.
=0 .Eo -o", 2.2 RELATIVE IMPORTANCE OF CATEGORIES OF .Q SMALL-SCALE INDUSTRY
- '- E .a' -An examination of the composition of the industrial activities under
= )-,- . taken in Sierra Leone can provide additional insights into the nature - = of the industrial sector. The number of establishments, the number of C employed, and value added are used in this monograph to portray the 18 Small-scale Industry in Employment Generation 19 distribution of industries by industrial category as well as by size of locality. The results of this distribution by employment and enter- a, , prises are presented in . " Table 2.3, while the industrial composition by .- value added is presented - in Table 2.4. "77 7 !. One of the most salient findings is the dominant position of tailoring activity in Sierra Leone's industrial sector. Tailoring ac- counted " ." for 31 per cent of the employment, 33 per cent of the establishments - - and 37 per cent of the value added within the small- scale industrial r4: sector. This finding is consistent with studies of . a ' small-scale industry in other African countries have and elsewhere, which also revealed that tailoring is the single most important small- E - scale industry.' 4 C, - Blacksmithing, which accounted for 12 per cent of the . - sector's value added, and carpentry, which accounted for 10 per cent, a - were the next most important industrial activities, but both followed tailoring by a large margin. 4 0 o The composition of industries, however, varid 4nportatly,.with -the size and:"lo6itbn - " of settlements* ii Sierra Leone. Although "0 " - - tailoring is ubiquitous in Sierra Leone, it is relatively more important N in - : 3 Freetown than in the smallest villages. Indeed, the more 'tra- .- ditional' crafts such as blacksmithing, weaving and mat making ' relatively more important are in the smallest villages, while the more " S . " 'modern' activities such as tailoring, vehicle repair and metal welding " are more important in Freetown.'- E . - These results not only reflect the importance of including location in the analysis of industry, but also provide support for distinguishing between 0 "' 'rural' and 'urban' small-scale industries. Although any division CD V 10 between rural and urban is arbitrary, 20 000 inhabitants has "44 - - been adopted as the dividing line in this study. Thus, Freetown, Koidu, Kenema, Bo and Makeni - localities with more than 20 000 . inhabitants - have been classified as urban, while all other localities . ,_. have been classified as rural. Using this classification scheme, 95 per a cent of the industrial establishments in Sierra Leone are - - rural located in .' localities and account for 86 per cent of the employment in the " 0 -S , ,...,:t 6 - industrial sector. Finally, 75 per cent of the value added by the small-scale than urban industries7 in Sierra Leone is generated in the areas. These results thus reflect the importancerural of rather rural industries and the need to include these enterprises in studies of the '0 industrial 4 t c S" M sector of Sierra Leone. Moreover, these results also point . '1 to the need to ensure that rural industries . . g0 are incorporated into < 8 Z 0 industrial studies in developing countries. < < (2
rq",'a'* V 'J0
TABLE 2.3 continued ISIC Typ f Localities Localities Localities Localities activity Localities Total Percentage less than 2000-5000 5000-2000 20 000-100 000 over 100 000 inhabitantsb (Bo, Kenema, Koidu, (Freetown) of Total Makeni) Estab. Empl. Estab. Erpt. Estab. Empt. Estab. Empl. Estab. Empl. Estab. Empt. Estab. Empl. 951 9512 RadioRepair services c c 10 9513 Vehicle 12 7 14 19 44 20 C c 20 98 61 56 131 0.1 0.2 9514 Watch 26 102 54 609 66 578 c c 10 16 15 166 1387 0.3 1.6 3909 Others 22 28 40 56 76 2 000 2 000 144 446 109 154 0.2 0.2 37 73 100 273 2 %63 Total 2490 3755 5.0 4.2 45 000 73 000 1 704 4 164 834 1995 1189 4368 1408 5039 50 135 88566 100.0 &InternationalStandard Industrial Classification. 100.0 bFinal estimaes number oased on projection from data obtained from the 24 sample of industrial establishments and industrial 'enumeration areas'. The estimate were obtained by multipying employment in each 'enumeration areas' by the actual figures reflecting the representation of that 'enumeration in that particular agricultural region. area' CIncluded in 'others'. SOURCE Data collected during Phase I of the small-scale industry component of the African Rural Employment Project. Njala University College.
TABLE 2.4 Sierra Leone: Value-added by small industrialcategoriesby location, 1974-75 Industrial Villages Localities Localities Locality Category less than Total Percent 2000-20 000 20 000-100 000 over 2000 (Bo, Koidu, 100,000 inhabitants Kenema, Makeni) (Freetown) Tailoring (leones) 2 836 000 534 000 571 000 Blacksmithing 1 982 000 4 923 000 37 427 000 77 000 11 000 Carpentry 385 000 9 000 1 524 000 12 Baking 558 000 219 000 122 000 (1) 259 1 284 000 10 Gara dyeing 000 175 000 264 000 698 (1) 345 000 5 Other 000 165 000 30 000 540 3 062 000 480 000 4 000 257 000 477 000 4 276 Total 000 32 77710000 2253000 1 398 000 1884000 13 245 000 100 (58%) (1/%) (11%) (14%) (1)Included with 'other'industries. SOURCE Data collected during phase 2 of the small-scale industry component of the African Rural Employment Project, Njala University College.
I-. 22 Small-scale Industry in Employment Generation Small-scale Industry in Sierra Leone 23 It is important now to examine in depth the economic parameters of the individual ties as well. Indeed, the data from the farm-level small-scale industrial firms in Sierra Leone. To study reveals, for ensure that example, that tailoring output accounts a reasonable approximation of the underlying population for on!y 36 per cent of the is provided, gross value of all outputs produced annual mean values of the input and output parameters by the village tailoring household; were farming output accounts for constructed from the randomly selected firms for the vast majority of the remainder. which data Correspondingly, the blacksmithing were available for the entire 12-month period. 9 Complete output accounts for 44 per cent 12-months' of the total output produced data were currently available for only 111 of the by the blacksmiths' households, while original 276 ran- carp.,ntry output accounts domly selected firms in 1974, and thus for some for only 17 per cent of the total output industries the produced by the 23 number of observations proved to be somewhat carpenters' households. Thus, the output geper low.' Preliminary ated by the village analyses, however, revealed that the mean value small-scale industries generally represents only of the parameters a portion of the did not vary significantly annual value of total output generated by these between those firms with complete data and those households. firms with incomplete data; this fact should consequently kept in mind be when interpreting the results presented in this section. 2.3.1 Seasonal variations in output 2.3 ANNUAL OUTPUT AND VALUE ADDED variationsThe annual value added and output data, however, in the level of activity over mask important The details of the year. Indeed, one the annual mean gross output and value added21 figures of the key for the major categories reasons that data were gathered for one year of small-scale industries in Sierra Leone in in the present study was 1974 are summarised to permit these seasonal variations to be more in Table 2.5 by location. The data reveal that precisely quantified. there were important The mean monthly variation in output variations in the mean output and value added of the major small-scale generated by industry groups located outside the villages the various categories of small-scale industries in Sierra is portrayed in Figure 2.1. Leone. In those The graph reveals that all the major industrial localities with 2000 or more inhabitants, for exam- groups were subject fo pie, the mean large monthly fluctuations in the level annual value added for baking (Le.2288), carpentry of activity. Indeed, for all the (Le.1625) and gara22 major industries, the mean output in the peak dyeing (Le.1567) was more than twice that of month was at least twice the mean output in the lowest month. tailoring (Le.739) and blacksmithing (Le.754). The output variations industry The blacksmithing were of a similar magnitude. exhibited the largest variation inoutput over the year, while u vtailoring The most exhibited striking result in Table 2.5 was that the mean annual the least. value added in The seasonal pattern of output, several industries appeared to vary importantly with however, is not the same for all the location. The major industrial groups. Tailoring - mean annual value added of tailors, for example, the most important of the small varied from scale industries is most active Le.203 in the rural villages to Le.1204 in Freetown: during the last quarter of the year carpentry and blacksmithing when the major harvesting activities are occurring. exhibited similar wide variations in this The seasonal peak parameter. Analysis for tailoring may be traceable not of variance procedures applied to these , only to the several important revealed di religious festivals and holidays that that the locational variation in the tailoring value occur during this period, but also figures added to the fact that the farmers, were significant at the one per cent level. Due primarily who are the major customers for tailoring to the products in the more rural localities, small number of observations, however, the locational earn much of their cash income the other variations in at this time. The blacksmithing industries' mean value added figures were not statistically and baking industries, on the other hand, attain their peak levels of activity in the significant. second quarter of the It should be noted that the low value added year. The seasonal and gross output which bread is peak for bread may be due to the fact that rice, for figures in the villages are traceable primarily a partial substitute, may be less available and rela to the fact that the tively more costly proprietors and their families are generally engaged during this period. The main explanation for the in farming activi- seasonal peak for blacksmithing activity is that the farmers, who are A TABLE 2.5 Sierra Leone: Annualmean output and value addedby majorsmall-scaleindustrialcategoriesandsize of localities, 1974-75 (leones) Industrial Variable Localities Category Less than 2000- 20 000- Over All those 2000 20 000 100 000 100 000 above 2000 Tailors Gross output 213.3 604.3 1 149.7 1 657.7 997.4 Material and service input 11.0 173.6 278.6 454.1 258.6 Value added 202.6 430.6 871.0 1 203.6 738.8 na 8 23 29 7 59 Gara Gross output - 3484.5 - 3484.5 Material and service input -- 1 917.5 - 1 917.5 Value added 1 567.0 - 1 567.0 n 2 Carpentry - 2 Gross output 78.0 3511.0 2 187.3 - 2452.0 Material and service input 7.1 2 241.0 473.0 - 827.2 Value added 70.0 1 270.0 1 713.5 - 1 n 624.8 6 1 4 - 5 Blacksmiths Gross output 245.5 1 361.3 706.0 1008.0 1 084.0 Material and service input 6.3 219.6 307.0 705.0 329.6 Value added 237.9 1 141.7 399.0 303.0 754.0 n 10 3 2 1 6 Baking Gross output - 5 015.5 - 5015.0 Material and service input -- 2 727.5 - 2 727.5 Value added 2 288.0 - 2 288.0 n 2 - 3
Goldsmiths Gross output - 600.0 1 591.5 767.0 958.5 Material and service input - 271.0 320.0 150.0 267.2 Value added - 329.3 1 271.5 617.0 691.3 n - 3 2 1 Watch repair 6 Gross output - - 1 249.5 185.0 894.7 Material and service input - - 259.5 41.0 186.7 Value added - - 990.0 144.0 708.0 n - - 2 1 3 Radio repair Gross output - 1 071.0 - 1 071.0 Material and service input - - 193.5 - 193.5 Value added - - 877.0 - 877.0 n - - 2 - 2 Shoe repair Gross output -- 508.0 1 547.0 - 1 027.5 Material and service input - 63.0 495.0 - 279.0 Value added - 445.0 1 052.0 - 7 485.0 n - 1 1 2 a number of observations; randomly selected firms only.
SOURCE Data were collected during Phase 2 of the small-scale industry and farm level components of the African Employment Project, Njala University College. 27 26 Small-scale Industry in Employment Generation Small-scale Industry in Sierra Leone in There are also major variations in small-scale industrial activity in the villages as well. Our survey reveals that the seasonal variation 22-- Blacksmiths in the larger localities. The data 210 Tailors the villages is larger than that found 210 eters from rural hinhouseholds t he Boliland area, for example, indicate that 200. of the households' time devoted to non-farm activity 190 I I the percentage varied from a low of 0.2 per cent in September, when the demand for 180I in March, when the farm labour is at its peak, to a high of 42 per2 4 cent 170" for farm labour is relatively low. 160- !demand 150 . 120140 - 2.4 MATERIAL AND SERVICES INPUT Zx 120 purchased material and services inputs I.11 The relative importance of industry, as an examination of Table 100 varies widely from industry to 90 - -"2.5 will reveal. These inputs, for example, comprise over 50 per cent large quantities of 60 "of the gross output value in gara dyeing, where cloth and dyes are purchased, and baking, where flour is the domi " I nant purchased material input. On the other hand, the purchased sI material and services inputs comprise less than 25 per cent of the 40..: ' . gross output value in tailoring, since tailors work primarily on cloth "brought to them by their customers. S- used by these 2 " The direct import content of the material inputs small-scale establishments also varies by industrial category. In the 0-4 industries, virtua!ly all the material inputs, 2 3 4 5 6 7 8 9 10 11 12 tailoring and gara dyeing JauaMONTHS December such as cloth, dyes, needles and buttons, must be imported. Thus, approximately 55 per cent of the gara dyers' and 20 per cent of the gross output values consist of imported inputs. 5 The material FIGURE 2.1 SierraLeone: Index of monthly outputof majorsmall-scale tailors' industries, on industries, 1974-75 (November = 100) input components of the other three major small-scale the other hand, are substantially lower and vary from 2 per cent of tailors and carpenters reflect the mean output value the gross output value of blacksmiths to 10 per cent for carpenters NOTE: -The indexes for produced flour, but this is for all localities while the indexes for blacksmiths and bakers are limited to and bakers. The bakers use domestically produced from imported wheat and they must import other ingre those localities with 20 000 - 100 000 inhabitants. blacksmiths use domestically produced charcoal and scrap SOURCE Survey data dients; the metal but import other metals; and carpenters use large quantities of timber but import plywood, the major demanders of blacksmithing services, engage in land domestically produced nails and sawn that the direct import clearing during the second quarter and thus require large quantities veneer and other woods. These results indicate above zero for all the major of blacksmiths' products and jervices at that time. In summary, these component of material inputs is clearly and in some cases is quite results thus reflect the importance of incorporating seasonal variation small-scale industries in Sierra Leone policies can have an into any analysis of small-scale industry and point to the potential substantial. Thus, in Sierra Leone, import industries. dangers of surveys limited' to one month. important direct impact on small-scale 28 Small-scale Industry in Employment Generation Small-scale Industry in Sierra ILeone 29 The small-scale industries in Sierra Leone, however, import a much smaller understated to the extent that firms rent share of their material inputs than do their large-scale their buildings and equip counterparts. According ment. The survey data, in fact, reveal that to data presented in the 1974 National the vast majority - 70 per Development cent of small-scale industrial establishments Plan, 45 per cent of the large-scale industries' i. Sierra Leone rent output value gross their premises. Moreover, approximately consisted of imported material inputs. The correspond- 36 per,cent of the tailors in ing figure for the the survey rented their sewing machines.2 9 Thus1, five major small-scale industries was 17 per cent. if the use of capital Thus, by small-scale industries is to be measured the small-scale enterprises in Sierra Leone require completely, it is impera less foreign rive to combine the rental exchange for material value of buildings and equipment with the inputs per unit of output than do their large- scale counterparts - a distinct advantage capital stock data. when foreign exchange is The combination of the stock and rental capital values, however, relativelyFinally, scarce. it should be noted that small-scale industries raises the important theoretical and empirical issue of what in Sitrra correct specification of is the Leone generally make very little use of purchased the capital input. Ideally, the capital services service inputs, rather than the Electricity, for example, is purchased capital stock figure should be used in the analysis by only 23 per cent of the the of tailors in those localities cost and production relationships of firms. outside the villages, and by a few bakers, Most studies, how blacksmiths and ever, use stock as a proxy for capital carpenters using more modern techniques of produc- services flows, but this stock tion. 6 Moreover, measure is valid only under the restrictive assumption only the 'modern' bakers purchase water. Thus, that the capital the vast majority stock components are of the same 3 of the small-scale establishments are not dependent durability and vintage. 0 Conse upon these services quently, in the present study, all the building inputs and therefore possess more locational and equipment capital stock variables initially have flexibility than do their large-scale counterparts. 27 been converted into capital services flows using the capital recovery formula:
2.5 CAPITAL SERVICES AND STOCK INPUTS 1 - (1+r)V Controversy and debate still surround the definition and measure- ment Where R is the constant annual of the capital input. Thus, the measures used capital service flow, V is the original to describe the (undepreciated) market capital input used in this study must first be value of the asset, r is the discount rate and n discussed. The magni- is the life expectancy tude, composition, sources and utilisation of the capital. This formula generates an annual of this capital input will capital service then be examined. flow from the capital stock data that is equivalent the to In this study, rental charge or 'user cost' of capital and reflects capital has been divided into three major categories: both deprecia working capital, tion and the opportunity cost equipment and buildings. Working capital is defined of the capital. to include cnly the The use of the capital recovery factor, however, value of the inventories of material inputs and requires two finished outputs held additional parameters: (1) the life expectancy by the enterprise.' These stocks have been of the capital stock (n) evaluated on the basis and (2) the discount rate reflecting the opportunity of the inventory levels at the beginning of the cost of capital (r). survey year. Equipment Data were collected on both the age and the expected is composed of the tools, machines and remaining lives furniture that of the equipment and buildings used are being utilised by the establishment. These by the small-scale industry assets establishme:,ts. To elininate have been initially valued at their original distortions arising from incorrect or purchase price on the basis inconsistent of information supplied responses supplied by individual establishments, by !he entrepreneur. The final capital cate- how ever, the mean expected life for gory is buildings and reflects the value of both each component )f the capital stock the building and land was 3 that are being utilised by the establishment. calculated and used in the anal Vsis. If only the value The appropriate discount rate is somewhat of those capital assets purchased by the enterprise more difficult to ascer are tain. Ideally, one would included in the analysis, however, the capital want to use the rate of discount that reflects figures will be the opportunity cost of capital in Sierra Leone but this rate is rather 30 Small-scale Industry in Employment Generation Small-scale Industry in Sierra Leone, 31 elusive. The capital market appears to 32 be highly 'fragmented' with The composition of the capital services also varies by location. artificially low interest rates in the commercial banking sector and More specifically, the building rental component is relatively more unduly high interest rates in the informal credit markets. The maxi- important in urban than in rural locations, ii~flecting the relative mum discount rate charged by the commercial banks in 1974, for premium on space in urban locations. In tailoring, for example, the example, was 12 per cent, while rates in the informal rural credit building rent accounts for almost 50 per cent of the capital services in markets generally exceeded 40 percent.3 3 In view of these considera- urban Freetown, but only 31 per cent in the rural villages. Thus, tions, it was assumed initially that a discount rate of 20 per cent location appears to influence both the amount and composition of the provided a reasonable approximation of the opportunity cost of capital services input and thus reinforces the importance of specifying capital in Sierra 3 Leone. ' This particular discount rate was, if any- location in any analyses of small-scale industry. thing, somewhat low and thus would the more capital-intensive firms. give the benefit of the doubt to Since capital data in other countries are almost always expresed Snecptldt nohrcutisaeams lasepesdi in the doecpitals of te annualfimenaj sock rather than in service flow terms, The details of the annual the capital figures for the categories of small-scale indlustriesmean capitalare summarised services usedin Table by the 2.6. major The major small-scale in stock terms. Theseindustries statistics, in Sierra arrayed Leone by industryhave also and been location, expressed are table reveals that there are wide variations in the mean annual capital presented in Table 2.7. services inputs. In those localities with 2000 or more inhabitants, for p nted in example, gara dyers use only Le. 18 of capital services annually, whileAn Table 2.7 blackmitsemp, e mia on fT bl 2. rvastht he e n c ptlso k blacksmiths empioya n averague o Le.1 of tsevins. annual values, an average of JLe.411 of these inputs. The annual den includingoL.80(27)frbakmtig working capital, ranged from Le.81 ($89) for gara dyeing to Le.1890 ($2079) for blacksmithing. Theh blacksmithinglcsihn capital services used by tailors, the largest group, falls between these figure, however, reflects the relatively high capital stock values of the extremesThe compcsition and amounts of theseto an capital average services of Le.121 per year. also varies from industry two 'modern' blacksmiths that entered into the random sample.?6 Teodu stry.In fthse cait ofsolervies20 mr iesr inm ntfry When the capital stock figures for Sierra Leone are compared to industry. In those localities of 2000 or more inhabitants, with example, the building rent ranges from for capital stock data generated for small-scale industries a high of Le.223 per year (54 Africa, the paucity of capital used by the Sierra Leoneelsewhere small-scale iJi per cent of total capital services) in blacksmithing, to a low of zero in industries becomes apparent. A study of small rural industries in gara dyeing, where all production is done outdoors. Working capital Kenya reports that the mean value of these firms' reported invest services, on the other hand, range from a high of Le.38 per year (25 ment in fixed capital was approximately $3245. Another study of per cent of total capital services) for carpenters, to a low of Le.5 (4 small-scale industry in Ghana reveals that firms employing one to per cent of total capital services) in tailoring, where very small levels nine workers possessed fixed assets of approximately $5000, while of materials or output inventories are retained. In general, working those with from 10 to 29 workers possessed approximately $64 000 in capital thus constitutes a relatively small 3 7 percentage of the total fixed assets. Thus, the capital stock possessed by the Sierra Leonc capital actually used by these small-scale firms. This result, however, small-scale industries would appear to be small even when compared does not necessarily imply that more working capital might not be with small industries elsewhere in Africa. needed in order to improve the economic eu~ciency of these firms.35 Finally. the results i Table 2.6 indicate that there are important variations in the mean annual capital services related to location in 2.6 INITIAL CAPITAL STOCK REQUIREMENTS Sierra Leone. In tailoring, for example, the annual capital services ranged from Le.39 in the rural villages to Le.296 in urban Freetown. These capital stock input data, however, do not reveal the nature of Carpentry and blacksmithing also exhibited similar wide variations the initial capital stock that was employed at the time these firms by location. Analysis of variance procedures applied to the data for were established and thus do not shed light on a potential barrier to these three industries, the only ones with sufficiently large numbers entry into small-scale industry. To analyse this aspect, initial capital of observations, revealed that the locational variations were signifi- stock data were obtained from the entrepreneurs during the detailed cant at the 5 per cent level. entrepreneur interviews. The initial capital stock figures relating to TABLE 2.6 Sierra Leone: Annualmean capitalservicesinput by majorsmall-scaleindustriulcategoriesandsize of localities,1974-75 (leones) Industrial Variable Localities Category Less than 200- 20 000- Over All those 2000 20 000 100 000 100 000 above 2000
Tailors' Building services 12 21 59 187 59 Equipment services 27 38 61 104 57 Working capital services - 3 7 5 5 Total capital services input 39 62 127 296 121 Gara Building services - - - - Equipment services - - 6 - 6 Working capital services - - 12 - 12 Total capital services input - - 18 - 18 Carpentry Building services 12 100 60 - 70 Equipment services 25 48 28 - 40 Working capital services - 20 44 - 38 Total capital services input 37 168 132 - 148 Blacksmiths Building services 12 72 366 240 223 Equipment services 26 132 168 288 178 Working capital services - 12 1 21 10 Total capital services input 38 216 535 549 411 Baking Building services - - 84 - 84 Equipment services - - 18 - 18 Working capital services - - 15 - 15 Total capital services input - - 117 - 117
Goldsmiths Building services - 20 60 120 50 Equipment services - 4 18 84 22 Working capital services - 0 42 0 14 Total capital services input - 24 120 204 86 Watch repair Building services - - 24 - 24 Equipment services - - 60 - 60 Working capital services - - - - - Total capital services input - - 84 - 84 Radio repair Building services - - 60 - 60 Equipment services - - 36 - 36 Working capital services - - 18 - 18 Total capital services input - - 171 - 171 Shoe repair Building services - 12 48 - 30 Equipment services - 0 12 - . 6 Working capital services - 0 24 - _42 Total capital services input - 12 84 - 48 aThe number of observations is the same as in Table 2.5.
SOURCE Survev data. 34 Small-scale Industry in Sierra Leone 35 C) n TABLE 2.8 Sierra Leone: Average initial capitalstock requirementsat time 8 4 0 'IT .D n 0 r- :of 00 " establishmentby major industrialcategoriesandsize of locality, 1974-75 (leones)1 : category'Industrial Localities ~ ~='0 C IIII - 00I0 I category , 4 Less2than 2000- 20j)00- Over All 2000 20 000 100 000 100 000 localities Tailors 68 45 62 50 Gara dyeing -- 31 31 56 . Z 'Z 0 'I ) o0CDM0 .T tn Carpenters 24 44 22- 47 c 10, 4 n 0 , ' ,I r- 0 aBlacksmiths*Baking 26 44 2)5 55 - - 90 90 - Others 22 72 65 152 63 41 " 'The number of observation,.is the same as in Table 2.5, except for the 2non-inclusion of the 'modein' blacksmith k ,-4r, 0, Ci - ,0 Data unavailable from firm. in Freetown due to lack of data.
C\ SOURCE Survey data. - the randomly selected enterprises are presented in T.ble 2.8, where , the data are arrayed by r 00 t r- M -DD The data industrial category and location. reveal that the initial capital stock requirements'or the i major small-scale industry categories were o-ite mod,;st. In ail the S.major industries, the mean initial investment required was Le.100. less than Indeed, gara dyers could enter with only Le.31 of initial capital and tailors needed but Le.56. These noted, are similar to the $75 initial capital results, it should be rualidres requirements reported for 3 rural industries in Kenya.te ' A . iiilyeureetparticularly striking result presented in Table 2.8, however, is the indication - "- -. -. that initial capital stocks owned do not necessarily decline .-- . "localities. as one moves to smaller, more rural The mean initial capital stock owned by tailors in the rural
E 2E = .- S, -: 0 ,. - -. village, for example, M' Le.50. One partial explanationis Le.68, while in urban Freetown it is only .2. •6 3 :2 - is that a sewing machine rental - 0':. m - . = "- €''U .: market has developed =. m in many urban areas, which allows urban 2i. tailors to rent machines. Almost all tailors in rural areas however, "Z • have to purchase machines.
. . ~,2.6.1 Sources of initial capital 0 - , "-c The sources of these initial funds are presented in Table 2.9, where -- , 2 0 the data are arranged by size of locality. The table reveals that 36 Small-scale Industry in SierralLeone 37
personal savings from agriculture, trade or business provided the majority (60 per cent) of the funds needed to establish small-scale industries in Sierra Leone, followed by gifts and family loans. 39 The paucity of funds obtained from either the Go~rrnment or the com mercial banking system provides a further indication of the rather C" fragmented and underdeveloped nature of the institutional capital market in Sierra Leone for small-scale industry.
2.6.2 Sources of capital for expansion -; The capital for the expansion of these establishments, on the other - - "hand, was generated from none of these sources. As Table indicates, 2.10 almost 90 per cent of the funds used for expansion were simply reinvested profits, reflecting perhaps both the high rates of -a -return to these C"C o"z particular activities and the fragmented nature of the capital market. One of the most striking results presented in Table .= -2.10 is the revelation that none of the establishments enumerated in .g Z- the sample had obtained funds for expansion from either the com cc mercial banking system or the Government. These Z results reflect the relatively nascent state of the capital market and the important ON C) played role by small-scale enterprises in generating and providing a ve ". - " hicle for investing personal savings in Sierra Leone.
_2.6.3 Excess capacity
Finally, it was important to ascertain if the capital stocks of the 0 small-scale industries were fully utilised.A Consequently, they are difficult to quantify although and subject to error, excess capacity 2
S -An examination excess capital capacityof Table existed 2.11 reveals that a substantial amount in all the major small-scale industries.of S=The excess capacity varied by major industry, however, ranging from .-0 i:_ :S anhigh of ingara dyeing.41 per Evencent in the blacksmithing industry to only 25 per more striking, however, was cent = E the indication that the amount of excess J", capacity varied by location. Li tailoring, example, there was 45 per cent excess capacity in the villages but onlyfor ... .U 0 24 per cent in Freetown. "0 0 found to exist in rural areasThe highest amount of excess capacity was may in the other major be traceable, at least in part, industries as well. This to demand consideration. These E.5 0. results thus indicated that, particularly a' 0 0 in the rural areas, the existing t9 capital stocks of small-scale industries in Sierra Leone were generally < aa not fully utilised. 0o0 0 :X Small-scale Industry in Sierra'Leone 41 40 Small-scale Industry in Employment Generation example. the mean yearly apprenticcship hours 2.7 LABOUR INPUT more inhabitants, for t of total worked varies from a high of 4942 in capentry (76 per cL to zero in gara dyeing.43 In tailoring, the The final input to be examined in detail is labour. In this section, the carpentry industry hours) worked by apprentices was 617, a magnitude and composition of the labour input will be examined, mean annual number of hours 39 per cent of the total tailoring hours. followed by a discussion of the characteristics of the two most number that represented exhibits similar wide variations by industrial important labour components, apprentices and proprietors. Hired labour also this labour component is used less The labour used by tile small-scale entrepreneur consists of his or category although in general apprentices. In those localities with 2000 or more her own or family labour, paid employees and apprentices. In terms extensively than annual hours of hired workers in tailoring is of the stock or number of workers, the apprentices represent the inhabitants, the mean total) while, in baking, the number largest component of the labour used by the firms in those !ocalities only 35 (2 per cent of the with 2000 or more inhabitants. Indeed, apprentices accounted for 42 increases to a mean yearly total of 3346 (62 per cent of the total). of the family labour input also varies per cent of the labour used, while paid employees and proprietors The relative importance to industry. Indeed, in the major industrial accounted for 17 and 41 per cent respectively, widely from industry labour is important only in gara dyeing and is The labour measure presented in the remainder of this section, categories, family categories. however, will be in terms of the flow of labour services. Specifically, negligible in the other industrial data reveal that the role of women in the major the labour services were measured in terms of the number of hours Moreover, these included in the survey is rather small. actually worked in production by each category of workers; conse- small-scale industrial categories be engaged actively only in the gara dyeing quently, the labour services variable did not include the hours of Women appear to 42 members contribute 16 per cent of the individuals when they were simply waiting for work. This method of industry, where women family labour hours. In all the othei. major measuring labour services was chosen because it not only produces industry's total mean annual however, the labour hour contribution of the most suitable input for production function analysis, but also small-scale industries, workers, proprietors or apprentices, is small. provides an indication of an establishment's excess capacity, women, either as family that the mean number of hours worked The details of the mean annual labour hours used by the major Finally, the data indicate the five major small-scale industries in those categories of small-scale industries in the various localities of Sierra by the proprietor in 2(X)0 is approximately 1000 and ranges from a Leone are ,,n,narised in Table 2.12. An examination of this table localities in excess of a high of 1300 hours. The propietors of those small reveals that, as with the capital services input, there are wide varia- low of 80) to located outside the villages thus appear to work about tions in the mean number of labour hours by the major industrial scale industries mean the s;,.me number of hours as farm proprietors." categories. In those localities in excess of 20(), for example. the and over-all magnitude of the labour services annual number ofttailoring hours worked was only 1572 hours, while in Soth the composition by location. The total mean labour hours in carpentry the mean annual number of hours worked was 6457. input, however, varied for example, is significantly lower in the The composition of this labour input also varied by industrial all the small-scale industries, category. This result is also ;i-parent from Table 2.12. where the total villages than in any of the other locations. Tailoring. where the total ranged from 474 in the villages to 1398 in Free hours of labour input by industry have been subdivided into the mean labour hours illustration of this variation. The low labour following categories: town, provides a good for small-scale industries in the villages, however, simply hour figure in rural areas. Indeed, an (1) proprietors' labour, refiects the part-time nature of this activity labour hours data (2) family labour (male, female, child), analysis of the combined farm and non-farm hours of those engaged (3) apprentices and iuaicatc that 66 per cent of the annual labour are devoted to farming (4) hired labour insmall-scale iiudustry activities in the villages non-village localities, on the other hand, the total The mean annual number of hours actually worked by apprentices activities. In the for carpenters and blacksmiths were largest in differs widely by industrial category. In those localities with 2000 or mean labour hours TABLE 2.12 SierraLeone: Annual mean labour hours by small-scale industrialcategory, labourtype andlocality (number ofhours) Industrial Variable' Category Localities Less than L clte 2000 20 000 Over 2000 20 All 000 100 000 !00 000 Tailors Proprietor labour localities Male family 474.0 869.0 labour 2 840.6 Female family 1 398.4 917.9 labour 2 ApprenticesChild family labour .6 .5 0 20022.3 .9 .1 0 Aprenticesur 0.9 Hired labour 323.5 Total 0 18.3 534.0 1 928.9 617.8 474.0 1 211.3 52.3 15.0 36.0 Gara 1 431.2 3342.3 1 Proprietor labour 3 342.3 1 572.3572.5 Male family labour - - 1 005.5 Female family labour - 1 .5 - 1 000.5 Child family labour - - 1.5 1.5 Apprentices Hired labour -15.5 Total - 15.5 Carpentry - 0 Proprietor labour 381.5 1.5 Mlfaiylbu2308.0 3 - 381.5 Male family labour 3 132.0 2 919.5004.5 - 2 - 12 362.0004.5 Female family labour Child fam. y labour 2 0 2 0 0 0 0 0 TotalApprentices Hired labour 0 2019.1502 15541.0 191.3 308 18 673.0 2292.5 153.0 3 403.3 - 6457.24942.2
Blacksmiths Proprietor labour 745 Male family 1 123.0 496.0 446.0 labour 2 801.0 Female family labour 0 0 2 0 - 0 Child family labour 0 0 0 Apprentices 2 0 0 2 0 0 Hired labour 691.7 631.5 775.0 Total 2 200.7 1185.5 745 669.0 - 323.3 3 015.3 1 796.5 1 221.0 Baking Proprietor 2 310.0 Male labour family labour - Female family labour 277.5 - 277.5 Child family - 754.0 labour - - 754.0 - 0 Apprentices - 0 -- 0 Goldsmiths I..'oprietorHired labour 0 Total labour - 0 - 4 171.0 0 Male family labour - 3 195.5 - - 43 171.0139.5 - 700.0 Female family labour 988.5 1 486.0 - 0 927.0 Child family labour 53.0 0 17.7 C- - 0 Apprentices 486.0 0 0 - 395.7 0 0 243.0 Hired labour 550.0 812.0 5.-1,.0 Total -0 - 0 0 1614.0 1 591.0 2298.0 0 Watch repair Proprietor labou r - 6 68.0 1 720.7 Male family 17.0 1 .0 labour - Female family labour 687.0 171.0 - - 515.0 Child family labour 0 0 0 - - Apprentices 0 0 0 Hired labour - 90 0 0 - 10 0 Total - 710 0 7.0.0 -:" 715.0 continued on page 44 171.0 534.0 44
.those Small-scale Industry in Sierra Leone 0 0 C 0 0D 2.7.1 Characteristics The of apprentices Nthe apprenticeship system provides labour for small-scale the primary vehicle for training industry. It is a system in which a young person serves a proprietor or master for a given period of order to learn a trade or craft. time in varies from industry The duration of the apprentieship to industry, as Table apprentice, for 2.13 reveals. A example, serves for only gara dyeing while a blacksmith's about a year and a half, apprentice serves for an average The service of of five years. a tailoring apprentice falls between averages three and a half years. these extremes and Ci The duration of the apprenticeship does not vary ,. importantly with location; indeed, analysis apprenticeshipprocedures indicated that of variance were not statisticallylocational significant. variations in the duration of I-Z.0 = Cc X 00 Cz= 0 0 . E= .proprietors A large number of CU " or the apprentices ai.e required z- . U C !2 during this period.masters a learning fee to pay their ° the Indeed, as indicated for the training o E . E- * small-scale industrial establ.shments by Table 2.14, 53they per receive - charged a fee for the trainingcent of ,0 = D. .-E_0 "a .- 0 given.thimalsc . The percentagel of firms requiring the fee in the major indus trial categories ranged from a low of 17 per cent in cent in gara dyeing. baking to 100 per CJ Whether or not a learning fee was required depended to a great 00. 2.". xtent on the location of the enterprise. In rural localities %an 2000 with less irms inhabitants, for example, approximately required a learning fee, while inFreetown 73 per cent of the z- 0 only 13 per cent of - C 0iffering he firms required underlying suchpattern a fee. of demandThis variation for and maysupply be oftraceable atnrentirc to a Small-scale Industry in Sierra Leotie 47 6 , TABLE 2.14 Sierra Leone: Percentageof small-scale industryproprietors Industry charging, learningfee to apprentices Localities, 0. i Less than 2000- 20000- C oi 2000 20000 20000 All Z 100000 localities .0 'u 0 Tailors n 79% 58% 56% n 10 33 0% 54% 10 33 34 8 00 Wo M-, ni Gara - 100% 85 "- n 100% 100% 100% - - 3 Carpenters 3 1 7 70% 69% n 75% 33% 68% (Z 10 13 Blacksmiths , 8 3 34 88% 50% 0% 0%z E Baking n- 17% -' - 8 4 6 0 0 - 0 18 Others 43% 81% •_ - 26% 20% 46% C14 _ n 7 Weighted 16 19 5 47 '0 " C grand mean 0 n = number 4 q 4 r-n- & of observations; based on all firms randomly and purposively selected that responded to this question. - . SOURCE Survey data. 6 ,-Since apprenticeship training is a basic recuirement for 0. many small-scale entry into .0 appears industries, the relative entry cost relating to training 00- to be higher in rural areas than in the urban areas.4" L IT 0 To ascertain the dimensions - " of the learning fee as a potential barrier, itisalso necessary entry . irather to examine the •o magnitude of the learning fee. As Table 2.13 reveals, similar for the major theindustrial learning fees in the rural villages are •- - /r oo , groups, amounting to approxi - mately Le.25 localities, the learningfor the feeduration of the apprenticeship. charged by those firms requiringIn theappears large to increase it O somewhat for all - . But, industrial categories except -cc given the predominance of the tailoring. learning fee does appear to learning fee in rural areas, the fali more heavily on the rural than 0 on the urban apprentices. C4LU •U, Finally, L. "0 the vast majority of the small-scale E received their own training industry proprietors the survey data indicate from the apprenticeship that 90 per cent of system. Indeed, -, the proprietors sampled 0M - had previously o served as apprentices. 6 0 Since proprietors generally 48 Small-scale Industry in Employment Generation Small-scale Industry in Sierra Leone TABLE 2.15 Sierra Leone: Percentageof 49 small-scale industrialproprietors' TABLE 2.16 Sierra fathers who were Leone: Percentageof small-scale industrialprcprietors farmers by industrialcategory and location with formal education by Industry Localities industr) and location Industry Localities Less than 2000- 20 000- Over All - z - O 2000 20 000 100 000 100 000 localities Less than 2000- Tailors 86% 71% 20 000- Over All n 61% 100% 72% 2000 20 000 100 000 100 000 localities 14 31 33 Tailoring 21% 15% Gara 6 841 35% 25% 24% Gr - 100% 84n14 34 n 33% 100% 71% 34 34 85%829% - 3 Gara - Carpenters 3 1 7 0% 25% 0%9 80% 83% 57% - 3 n 33% 72% Carpenters 4 13% 10 12 7 3 10% 25% 38% nBlacksmiths 75% 32 n 33% 24% 8 38%8 25%4 - 10 12 8 - 2050% n10Blacksmiths 3 33 0% 80% 25%4 Bakers - 75% -- 225% 80% 0% 50% n - Baking - 8 5 7 20 10% 29% 71% Others 25% 53% n - 33% 39% 100% 48% 10 7 7 n 8 17 Others 0% 24 Wegtdn 18 5 48 12% 53% 20% 2 n 9 17 Weighted520 9 17 1919 5 50 26% grand mean 63% grandWeighted mean n = number of observations; based on all randomly and purposively selected 23% firms that responded to this n = number of observations; based question. on all randomly and purposively selected firms that responded to this question. SOURCE Survey data. SOURCE Survey data. require some technical skills before they can launch a business, the apprenticeship system thus Nigerian proprietors in the plays an important role in the generation surveys of both Callaway and Harris.48 of new small-scale enterprises. The One might hypothesise that the other characteristics of the pro- different results were traceable, at prietors must now be examined least in part, to the absence in more detail. of rural proprietors from the Nigerian surveys. Table 2.15, for example, does indicate some tendency for the percentage rural localities. ofVarious farmers to increase as one moves 2.7.2 Characteristis of proprietors Chi-square tests, however, to the smaller except for indicated that, significant bakers, these locational differences were not statistically There is a wide range of socio-economic at even the 10 characteristics that might per cent level. shed direct light on the proprietor's Another potentially important ability to operate the firm. The characteristic of the proprietor is occupation of the proprietor's father, the level of formal education. for example, could provide an One might hypothesise that formal indication not only of the education contributes to inter-generational occupational mobility, managerial, organisational and technical but also the psychological attitude skills. The extent of the formal education 47 of the proprietor -owards operat- of proprietors is provided ing the firm. In view of by Table 2.16, where Sierra Leone's pervasive agricultural the percentage of proprietors with is perhaps not base, it education is arrayed any formal surprising that farming was the primary by industry and location. two thirds occupation of result The most striking of the fathers of the industrial proprietors. is the low percentage of proprietors The two-thirds - with any formal education figure for Sierra Leone is higher, 77 per cent possessed no formal education however, than that reported for at all, a high percentage even by African standards. In Callaway's sample of proprietors in 50 Small-sca.le Industry in Employment Generation Small-scale Industry in Siehra LeL,,, 51 TABLE 2.17 Sierra Leone: Percentageofsmall-scale industrialproprietors who keep TABLE 2.18 SierraLeone: Age ofsmall-scale records. Industry industrialestablishments,1974 Age of business.(years), Industry Localities LocalitiA Less than 2000- 20 000- Over All Less than 2000 20 000 100 000 100 000 localities 2000- 20000- Over All 2000 20000 1000 Tailoring 0 100000 localities 0% 3% 21% 0% 9% n 14 Tailors 14.4 11.3 33 34 8 89 n 12.3 7.3 11.8 Gara - 67% 50% 14 35 33 n 0% 50% Gara 8 90 3 4 - 16.7 19.0 13.0 Carpenters 1 8 n 17.2 10% 15% 38% 33% 3 3 1 7 n 21% Carpenters 14.9 10 13 8 3 34 11.5 10.5 18.0 12.8 Blacksmiths 0% n 10 13 0% 25% 5% 8 3 34 n 9 8 Blacksmiths 14.0 19.7 4 - 21 n 14.0 16.0 Baking - 0% 57% 9 7 4 - n 86% 46% Baking 20 - 8 7 7 - 7.8 6.4 6.9 Others 22 n - 7.1 0% 6% 32% 20% 16% 8 7 7 22 n 9 Other 19.3 17.4 17 19 5 50 n 10.8 12.4 14.6 Weighted 8 16 19 5 grand mean 48 17% grandWeighted mean n = number of observations; 12.6 based on all randomly and purposively selected n = number of observations; based on all randomiy firms that responded to this question, and purposively selected firms that responded to this question. SOURCE Survey data. SOURCE Survey data. Ibadan, for example, only 24 per cent had no formal education, while Harris reported that, Finally, the number of years that proprietors' among his Nigerian proprietors, only 13 per establishments have cent had no formal 9 been operating should also provide an education. Although there was some variation indication of his or her ability in the level of formal to run the firm. The data on the number education by industry and location, except of years since the proprie baking, tors' businesses were established none of these variations were statistically significant. are summarised in Table 2.18, Clearly, wh2re they are arranged by as noted previously, non-formal education apparently location and industry. The mean number provided the of years that these major source of training for small-scale small-scale establishments have been operating industrial proprietors. Sierra in Another potential Leone is approximately 13 years. The age of indicator of the ability of the proprietor to firms appeared to operate the firm vary by location and industry, with the is whether or not the firm keeps records or business urban tailoring and baking accounts. As the data establishments tending to be younger than their in Table 2.17 indicate, only 17 per cent of the counterparts else firms kept where. However, analysis of variance procedures even a very rudimentary set of business accounts applied to these or data indicated that these age variations records. The extent to which these business by location and industry were did practices were utilised not statistically appear to vary somewhat by industry and location. significant. It was more Several of the socio-economic prevalent, for example, in the larger urban characteristics of the proprietor that locations and in the have been examined baking industry. On the basis of Chi-squared do, however, vary importantly by establish tests, however, the ment, and variation was found to be statistically many of the characteristics have been hypothesised significant only for bakers and ifluence to the performance of the firms. These hypothesised tailors. relation slips will be tested and reported in Chapter 3. Small-scale Industry Output and Erhployment' 53 the demand for these products will decline absolutely as the level of rural incomes increases.' It is thus imperative & -t the various compo nents of the demand for the products of small-sc.% 'ndustry in Sierra Leone be analysed. There are three major sources 3Determinants of demand for the products of of Dem and small-scale industry in Sierra Leone. The primary source is that demand generated from the incomes of rural ahid urban consumers. for and Supply of A second mall-scale Idu source of demand arises from the backward and forward Sml-c l ndustry production linkages with the agricultural and large-scale industrial t al sectors. The final important source of demand for the products of ismall-scale industry is that provided by the foreign or export sector. Ou put and Emp oym ent in Each of these Sierra Leone demand sources will now be examined. The income elasticity of demand is one of the crucial parameters required for analysing the linkage between rural and urban incomes, and the quantitv of small-scale industry products demanded. Indeed, both the magnitude and sign of the income The major determinants of the demand for and supply of out and elasticity of demand coefficients for small-scale industrial products are central figures in employment generated by the small-scale industrial sector oi iierra the debate over the future role of these activities. Hymer and Res Leone will be analysed in this chapter. The primary factors influenc- nick, for example, argue forcefully that, in the rural areas, the ing the demand for the goods and services of this sector will be products of local small-scale industries (Z-goods) are 'inferior' analysed first. Attention will then be focused on the supply factors goods and thus that the demand for and production of these goods influencing small-scale industry. Since the full spectrum of produc- will decline as rural incomes rise. In addition, these particular income tion choices and firm types needs t. be examined, data from both the elasticity parameters are important for tracing and measuring the randomly and purposively selected firms in our 1974 survey will be indirect effects on small-scale industry output and employment of included in the supply analyses undertaken. The production proces- policies or projects designed to influence other sectors. There ses within each of these industries will then be examined have and their been, however, virtually no empirical studies designed relative factor proportions determined. Variations to verify the in these factor magnitude and sign of these important elasticities. proportions by location and firm size will next be considered. The Fortunately, the rural consumption component2 of the overall returns to the proprietor and the economic profitability of the major Sierra Leone project has undertaken the required rural household small-scale industries can then be estimated and analyseo. The chap- income and expenditure surveys to permit the income elasticities ter will conclude with an examination of those entrepreneurial for charac- locally produced small-scale industry commodities to teristics that appear to be associated with be estimated. the economic profitability Specifically, the survey was designed of these establishments. to obtain a detailed breakdown of household expenditures on individual non-food items by origin of production. Thus, it was possible to distinguish between those com modities 3.1 DEMAND produced by small-scale industry and those produced by large-scale firms or imported from abroad. Although a wide range of functional Demand considerations are important in determining the economic forms and independent vari ables were used in estimating the relevant elasticities, the following viability of small-scale industries. Several scholars have argued that semi-log model was considered to be most reasonable: the demand for the products of small-scale industries is severely limited and indeed have concluded that, at least in the rural areas, C, = A + b, In Y, + b2 In N, + b3 In S, + e, N5 52 54 Small-scale Industry in Employment Generation Small-scale Industry Output and Emplyment 55 TABLE 3.1 Sierra Leone: Total expenditure elasticities of demand by rural results challenge the rather widely held contention householdsfor major small industry products, that the products 1974-75 of small-scale industry in rural areas are. all 'inferior' commodities. industry b, Expenditure elasticity Indeed, they indicate that the demand for these products should at mean level of generally be expected to increase as 5 total rurl incomes increase. Thus, expenditure' rather than being viewed as an overriding constraint, the demand 1. Sir" ' cale tailoring 1.76 .72 induced from rising incomes should be viewed as a positive force for prodLcts ( .57)2 the growth of small-scale industry in Sierra Leone, at least in the near 2. Small-scale gara dyeing 3.76 .71 6 products future. ( 1.20) The demand for small-scale 3. Small-scale blacksmithing industries stemming from the back 1.16 .92 ward and forward linkages with the consumption products (not ( .76) agricultural and large-scale industrial sectors, on the other hand, must be obtained from the including farm inputs) ind us tr ia se to rs, o the a ious e o f th e 4. Small-scale carpentry 3.22 1.35 relevant input-output products coefficients of the various sectors of the Sierra ( 1.04) Leone economy. The required 5. Small-scale baking .49 input and output data relating to the .61 agricultural sector were generated by the farm-level component products ( .34) 7 of the project, while the data relating to the large-scale industry 6. Other small industry 2.80 .83 were obtained from unpublished surveys conducted by the Central Plan products ( 1.30) o in Uendtt 7. All small-scale industry 13.19 .85 ning Unit. products ( 2.41) An examination of the relevant data, however, reveals that the 'Expenditure elasticity = backward and forward linkages of the small-scale industrial b, not yet very extensive. The linkages sector are 2Standard with the large-scale sector, for errors are in parenthesis. example, are limited solely to the backward linkage from the small SOURCE Byerlee and King, 'Factor intensities and local linkages of rural scale bakers to 8 consumption the large-scale flour mill. There are few'forviard patterns in Sierra Leone' in American Journal of linkages yet established from Agricultural the small to the large-scale sector and Economics, Vol. 60, pp. 197-206. Computations thus the large-scale sector have been batsed on a random sample of 21)3 households. provides few sources of intermediate demand for the products of the small-scale industrial sector.9 The general paucity of intra-industry where C. is 'he total consumption expenditure linkages no doubt is related import on each category of antly to the low level of industrialisation in Sierra Leone. small industry commodity purchased by the household, Y is the total The linkages of the small-scale industrial sector with agricultural consumption expenditure by the household, N is the household size, sectors axe, however, more extensive than those with the large-scale and S is the subsistence ratio of the household.3 industrial sector. The backward linkage from rice milling to rice The expenditure elasticities at the mean level of tota! expenditure production, for example, is very strong in Sierra Leone." Palm and for the major commodities produced by small-scale industry in Sierra cocoa-processing activities are of minor importance, although these Leone are summarised in Table 3.1. The coefficient b,, upon which have not been examined in the Sierra Leone study. the expenditure elasticities were based, was positive and significantly The forward linkage from small-scale industry to the agriculture different from zero at the 0.5 per cent level for all regressions, except sector in Sierra Leone is limited primarily to blacksmithing. As those for baking and blacksmithing. Table 2.4 reveals, over 93 per cent of the blacksmithing value added These expenditure elasticity figures, particularly the elasticity of is generated in villages, while approximately 28 per cent of the 0.85 for all small industry products, indicate that, at least in rural blacksmithing output in those localities with 2000 or more inhabitants areas, these goods were clearly not 'inferior' .4 The elasticity coef- is also destined for the villages." Moreover, a preliminary analysis ficients for the individual major small of industry categories products the data from the localities with 2000 to 20 000 inhabitants indicates were positive ranging from 0.6 in baking to 1.35 in carpentry. These that approximately 90 per cent of the blacksmith's output is some 56 Small-scale Industry in Employment Generation Small-scale Industry Output and Employment 57 form of a farming input. The primary blacksmithing activities fo where which the agricultural production occurred. The market for the outputs of gara sector provides an intermediate demand an dyers, farm tools, blacksmiths and carpenters, on the other hand, is somewhat such as matchets, hoes, knives and axes, and the repair o 8 farm less localised.' These data indicate, for eVample, tools and equipment. Indeed, the data generated by the farm that 37 per cent of the gara cloth, 32 per cent of the blacksmiths' output level study reveal that the average Sierra Leone farm household and 15 per cent it of the carpenters' output flow outside the locality 1974-5 purchased approximately Le.6 12 -where the com of farm tools. Of these farrr modities were produced. tools, approximately 90 per cent were domestically 3 produced b) blacksmiths. With the average farm household in 1974-5 producinj approximately Le.481 of agricultural output, approximately Le.1 ol 3.2 SUPPLY domestic blacksmithing output was demanded for every agricultural output.' 4 Thus, Le.100 ol the agricultural sector provided an im- It is now portant source of intermediate necessary to examine the major factors influencing the demand for the products of the supply of blacksmithing sector, small-scale industrial commodities. Specifically, the array of alternative production processes within each of the The final source of demand for the products of the small-scale major small industrial scale industry groups will be scrutinised ard sector is the export or foreign market. Huddle and Ho,'" their respective factor for proportion ratios will be calculated. The efiects of location example, have argued that the international demand for tradi- and firm size on factor proportions will also be examined. The tional goo:s produced by small-scale industries is quite high. Indeed, returns to the their proprietor and the economic profits earned by the various study indicated that the income elasticity of demand in high- small-scale industry groups will then be analysed. An analysis income countries for a broad grcup of culturally oriented products of the factors influencing the supply of entrepreneurship concludes this chapter. was greater than one. The data fenerated by the small-scale industry survey indicate that there is inc'eed mand an important international de for the products of one Sier:a Leone small-scale industry - gara 3.2.1 Alternative production processes and their factor proportions dyeing. The 'destination of sales' data of gara dyeing firms revealed that approximately 18 per cent of tle production of that industry was A process, according to Dorfman,' 9 exported. Thus, failure to include the export 'is a ,pecific method for per market in an analysis of forming an economic task.' the demand for the products of small-scale Each process is assumed to use the input industry may understate factors in fixed proportion, the existing and potential market size. and thus factors can be substituted for one another only if The analysis of the three major processes can be substituted. If, as in a fixed sources of demand for the products coefficient Leontief of small-scale industry has thus indicated production function, only one process is assumed that the existing and poten- to exist in an industry, tial market is clearly much 6 no process or factor substitution is possible. stronger than Hymer and Resnick On the contended and that the inter-sectoral other hand, if infinite numbers of processes are assumed to linkages and indirect effects are exist as in the neoclassical not insubstantial. Thus, one should formulations of production, continuous not necessarily presume that the process and demand for and production factor substitution is possible. To the extent that the of these industries in Sierra Leone has or actual number will necessarily decline.' 7 of processes utilised by any small-scale industry falls between Finally, the analysis of the these extremes, an examination of specific processes can 'destination of sales' data indicates that provide useful there are important variations additional insights into the nature of the production in the locational pattern of the demand relationships. within Sierra Leone The processes or techniques of production utilised by for the major commodities produced by the the small-scale major small-scale industries in Sierra Leone must thus be exam industries. The demand for products of the tailoring and baking ined to ascertain the degree -if process choice currently industries, for example, are very localised with ,ery little of available. the An investigation of the processes being utilised product flowing to other localities. Indced, 95 per cent of the in the major small-scale industrial groups in Sierra Leone reveals that bread and 97 per cent of the tailoring outputs remain in the locality more than one, though not an infinite number of processes, can be delineated 58 Small-scale Industry in Employment Generation Small-scale Industry Output and Employment 59 within each industry. To provide a sharper focus on the issue of extreme, there are several more process choice, however, only 'modern' bakers who produce bread the two major processes within each in large electric 2 ovens and who also use automatic mixers, industrial group will be discussed in this chapter. 2 kneaders " For simplicity, the and rollers. It should be noted, processes will be classified into two groups however, that in between the two - 'traditional' or 'modern extremes, there are although other processes being used which will be elabor the terms do not necessarily accurately portray the tech- ated nique used. upon in Chapter 7. 1 Having delineated the In the tailoring industry, two major process options within each the major process distinction arises be- industry, tween it is next important to ascertain whether, in fact, the factor those more 'traditional' proprietors that engage in only ordi- proportions for each of these major processes differed. nary sewing and tailoring, producing such products as common These factor shirts proportions are generally measured in terms and dresses, and those more 'modern' proprietors of the average ratios who also sew that relate capital, labour and output. embroidery designs on fabric. Thus, to determine the extent Although both groups use sewing of factor proportion machines, those tailors variations, it was necessary to compute the that embroider require a much more sophisti- output-capital, output-labour, and labour-capital ratios for each pro cated and more expensive type of machine than those who engage ordinary in cess within each industry. work. It should also be noted that even within each of these In addition to delineating processes, however, major categories there are further process subdivisions, these three econ which will be omic ratios provide useful discussed in detail in Chapter 5. insights into the production relationships. Firstly, they provide one measure of the factor In the gara dyeing industry, on the other intensity of produc hand, a :naiii process tion in the various industries. 3 distinction centres on the type of dyes used. Secondly, they indicate the relative The majority of estab- productivities of 2 lishments use imported, the factors of production. ' Finally, they provide a synthetic dyes in the manufacturing of the clue as to the amount of output and employment generated per gara cloth. There are still, however, a few, unit more 'traditional' estab- of the scarce capital input.2 lishments that continue to use the native indigo dye in the manufac- turing The mean values of the output-capital, output-labour, process. It should also be noted that there are further process and labour differences capital ratios of the two process options for the major small-scale related to the organisation of production and the type of industry categories of Sierra Leone are presented cloth used for dying, process differences that in Table 3.2. The will be discussed in output variable has been measured detail in Chapter 6. in terms of value added rather than gross output. Moreover, The major process distinction the labour and capital inputs in this table have within the carpentry industry centres been measured on the type of tools and equipment in terms of the flow of services rather than stocks 2 6 used. The great majority of more An examination 'traditional' of Table 3.2 reveals that the factor combinations proprietors use only simple hand tools such as hammers do generally appear to vary depending on type and saws. There are, however, a few more 'modern' of process utilised. firms that Indeed, analysis of variance procedures employ rather sophisticated machines such as applied to these data indi electiic saws, sanders, cated that the mean value and planes. 2' of the economic ratios that reflect these differences Within the blacksmithig in factor combinations were statistically different from industry, the process distinction also one centres another at the 5 per cent significance level. Each of the ratios on the type of equipment used. The vast majority of black- smiL,:. relating to these processes must now be examined. use the more 'traditional' tools and equipment such as small, hand-operated The output-capital ratio provides not only a useful measure bellows and small hammers. On the other hand, there of average capital productivity, but also an indication of the are some more 'modern' blacksmith firms, however, that use capital electric intensity of productior,. Moreover, if one forges and drilling machines as well as welding assumes that capital is the equipment. only scarce factor of production, Finally, within the maximum output is obtained when baking industry the main distinction between 2 processes the output-capital ratio is maximised. 7 Given the relative paucity centres principally on the type of oven used. There are, for of example, capital in Sierra Leone, a high output-capital ratio would thus a large number of the more 'traditional' bakeries that bake pro vide a good indication that a process or industry bread in large mud ovens costing was using this approximately Le.50. At the other resource effectively. 60 Small-scale Industry in Employment Generation Small-scale Industry Output and Employment 61 TABLE 3.2 Sierra Leone: Mean values of the output-capital,output-labour, and labour-capitalratios output-capital ratios. by process for major small-scale industry The difference between the processes highest and lowest output-labour with categories, 19745 ratios. was less than five times, whereas the corresponding difference between the processes with the Industrial categories' Output-capital Output- Labour- labour capital highest and lowest output-capital ratios was over one hundred (Le per houri The output-labour ratio or times. (Hoursper Le) is generally higher average labour productivity, however, Tailors for the more 'modern' process within 'Traditional try. Indeed, each indus 7.0 0.46 15.2 only in blacksmithing and gara does the average productivity for the 'traditional' labour 'Modern' exceed that of the 'modern' process. 6.2 0.55 11.3 The 'modern' blacksmith possessed Gara dyers the lowest average labour pro 'Traditional' ductivity 68.0 0.86 among all the processes of the major 'Modern' 41.0 78.8 'modern' industries, while 0.97 42.3 bakers and carpenters possessed the highest. The labour-capital ratio provides Carpenters another useful measure of the 'Traditional' 29.6 0.61 factor proportions and, more specifically, 'Modern' 48.5 intensity an indication of the labour 3.0 0.98 of production and the amount of employment Blacksmiths 3.1 unit of capital. Given generated per the relative capital scarcity and of abundance 'Traditional' labour in Sierra Leone, 11.9 0.36 processes and industries with higher 'Modern' 33.9 capital labour 0.5 0.25 ratios would generally appear, at least 2.0 making the on a crude level, to be Baking most effective use of the factors of production. 'Traditional' Table 3.2 reveals 19.0 0.5.9 38.0 that the labour-capital ratio, or labour-intensity, 'Modern' 5.0 is always highest for 1.18 4.2 the more 'traditional' process within industry. The differences in the factor each 'Includes both randomly intensity between processes in and purposively sampled firms in those localities carpentry, blacksmithing and baking with 2000 or more inhabitants. were particularly large. -.. SOURCE Survey data. The labour-capital ratio also varies widely by industry. For 'tradi tional' processes, it ranges from a high of 79 in gara dyeing to 15 in tailoring: for 'modern' processes, it One of the striking results ranges from 42 in gara dyeing to 2 evident from Table 3.2 is the indication in blacksmithing. These results, it that the output generated per those generated by the output-capitalshould be noted, are similar to unit of the scarce factor - capital - is data. highest for the more 'traditional' Indeed, these labour-capital processes within each major indus- and the output-capital ratios shed try. Indeed, the average light on the much discussed potential productivity of capital in the more 'tradi- conflict between output and tional' processes within empioyment goals.' Such a carpentry, blacksmithing, and baking conflict could supposedly arise, for industries was at least three times example, if processes or industries that of the more 'modern' proces ies. with relatively high labour-capital There are also extensive variations ratios (i.e., relatively labour intensive) in the output-capital ratio by also possess relatively low industry. For 'traditional' output-capital ratios. However, processes, for example, the output-capital if the relative magnitude of a labour ranges from a high of 68 capital and an output-capital in gara dyeing to a low of 7.0 in tailoring, ratio are of the same order, then the while for 'modern' potential conflict disappears and output processes, it ranges from 41 in gara dyeing and employment can, at least in blacksmithing. Thus, to 0.5 in a static sense, 2 9 'traditional' gara dyets possessed the highest be jointly maximised. average If one compares the capital producLivity while 'modern' blacksmiths output-capital and labour-capital ratios appeared to marised sum in Table 3.2 there is no apparent conflict possess the lowest. between employ The output-labour ratios, on the other hand, do not exhibit the ment and output within the same amount of variation Leone. 'Traditional' gara dyeing,small-scale industry sector in Sierra between processes or industries as do with the highest output-capital the ratio, also possesses the highest labour-capital ratio, while 'modern' 62 Small-scale Industry in Employment Generation blacksmithing, Small-scale Industry Output and Employment with the lowest output-capital ratio, also 63 possesses the TABLE 3.3 Sierra Leone: lowest labour-capital ratio. Indeed, if Output-capital, output-labour, labour-capital one ranks all the major small- ratios scale industry for 'traditional' small-scale industries by industry processes in Sierra Leone in terms of both and location, their 1974-75a output-capital and labour-capital ratios, the two rank orderings turn out to be exactly the same. The processes that produce the highest Industrial average outputs per capital Variable' Localities unit, - 'traditional' and 'modern' gara dyeing, 'traditional' carpentry, Less, 2000- and 'traditional' baking - also gen- 20 000- Over erate than 20 000 250 000 the largest amount of employment per unit of that 250 000 is the apparent scarce factor capital, which of production.' Consequently, there is 2000 no conflict, at Tailors Output-capital ratiob 5.3 least in a static sense, between output and employment 7.4 8.4 2.3 in the choice of processes Output-labour ratio .4 .4 either within or between industries in .5 .3 Sierra Leone. This Labour-capital ratio 13.3 result demonstrates the importance of incorpor- 22.1 16.0 7.6 ating these Gara Output-capital ratio significant process options into analyses - - 68.0 of small-scale Output-labour industry. ratio - - .8 Labour-capital ratio 3.2.2 Variations in factor proportions - - 78.8 - by location Carpentry Output-capital ratio 1.9 38.0 Output-laboar ratio 30.9 9.2 .2 .6 .8 The foregoing discussion of processes Labour-capital .2 and factor proportions, how- Blacksmiths ratio 8.3 94.9 52.4 33.6 ever, has been devoid of Output-capital ratio 6.3 13.7 locational considerations. Consequently, it 6.4 - is now important to Output-labour ratio examine whether or not there are any locational .3 .4 .3 variations in the Labour-capital ratio 21.0 factor propok-tions of these small-scale industries. 48.8 22.1 - one hypothesises If Baking Output-capital that capital is relatively more expensive ratio - 31.9 10.5 is and labour Output-labour relatively less expensive in more rural locations, ratio - .3 .5 " then the rural Labour-capital industries would be expected to be ratio - 85.0 35.6 more labour-intensive than their 'Includes urban counterparts. both randomly and purposively sampled 'traditional' To shed light on this issue, therefore, the relevant firms. 3 bOutput is measured in terms of economic ratios ' for the 'traditional' processes value-added. within each industry SOURCE have been computed Survey data. by location and are presented in Table 3.3. An examination of this table indicates that for many of the 'traditional' processes the factor proportions did appear to vary by location, together Indeed, as the size of the locality declines and both analysis of variance procedures applied to reach their peaks that the locational these data reveal in the 2000 to 20 variations in the labour-capital and output-capital 000 size range. ratios In view of this tendency for tailors, carpenters, and blacksmiths are for these two ratios to increase as the size significant at the 10 of localities declines, per cent level. The labour-capital ratios the relatively low labour-capital and output in these industries, for capital ratios example, increase as the size of the in the villages would appear to be anomalies. locality declines and reach their ratios, These low maximum in those localities however, are due largely to the part-time nature with from 2000 to 20 000 inhabitants. In of small-scale the villages with fewer industrial activity in these areas. As a result, than 2,00 inhabitants, the labour-capital ratios capital is idle for long are generally below periods and not fully utilised. It is of interest those in the larger localities. What is even to note, however, that striking, more blacksmithing activity, which however, is the discovery that the output-capital in other areas possesses the lowest ratios for economic ratios, these industries follow the same locational possesses the highest output-capital pattern as the labour- capital and labour capital ratios. Thus, the labour-intensity ratios in the villages. These results thus indicate and capital productivity variations locational for these 'traditional' in factor proportions, are important and should processes of small-scale industries do increase be incor porated into analyses of small-scale industry. 64 Small-scale Industry in Employment Generation Small-scale Industry Output and Employment 65 3.2.3 Variations in factor proportions by firm size taken elsewhere in Africa. In Steel's study qf small-scale industry in Ghana for example, Itwould also those firms employing no 'wage workers' be of interest to ascertain how the factor proportions f( were reported to possess an original cost of c~pital scalethese industriessmall-scale industrial processes compared with per worker ratio of in Sierra Leone as well as those for boththose for larg $435, while those firms with large- an $1,218 of capital per worker.from one to nine wage workers possessed small-scale industries elsewhere in Africa Child, in a study of a sample of small and other developin rural37 industries areas Since most countries and international in Kenya, found that the capital (stock)-labour agencies express thei for these ratio factor proportions in firms was $795. Thus, even when compared terms of the stock rather than the service flow with small-scale of labour and capital, industries elsewhere in Africa, the labour intensity the Sierra Leone ratios for small-scale industr of Sierra Leone's have also been expressed small-scale industry would appear to be relatively into stock terms to facilitate these interna high. 32 In addition to the labour-intensity tional comparisons. comparison between large- and The small-scale industries, it is first issue to be examined is whether the small-scale of interest to ascertain whether or not the industry i output per unit of capital more labour-intensive than its large-scale counterpart stock for small-sc-le industries is higher within Sierr; than that for the Leone. In this exercise, capital stock large-scale enterprises. Fo the major small-scale per worker rather than it! industries reciprocal will be used in Sierra Leone, the output (i.e., v' as a measure of labour intensity, since tht ie added) capital former measure is most stock ratio is estimated from the survey data widely used internationally. The capital stoc to be 2.1. The compar per worker for all able ratio for the twenty eight large-scale industries the major small-scale industries in Sierra Leone in Sierra Leone, was Le.380 ($418) and on the other hand, is no higher than 0.838 ranged from Le.15 ($16) per worker foi Thus, small-scale industries 'traditional' generate o,,er twice the amount of output gara dyeing to Le.2,504 ($2,754) per worker per unit of the scarce factor 3 3 for 'mod- capital than do the ern' bakeries. 'Traditional' tailors, the larger firms. The evidence from those few coun largest and most ubiquitous tries with group, possessed a capital stock per similar data also indicates that smaller firms worker of Le.197 (S217). The produce generally comparable figure more output per unit of capital than larger for the 28 'large-scale' firms with over 50 workers firms, although in Sierra Leone, several exceptions appear to exist.39 Since on the other hand, was Le.6.645 ($7,310) however the small firms in 34 per Sierra Leone, have also been worker. Thus, by this measure of labour intensity, shown to be more labour intensive than the small-scale the larger firms, the firms are approximately 17 times more labour-intensive apparent conflict betwcea output and employ than the ment once again dissolved. 4 larger firms in Sierra Leone. ' Small firms in Sierra Leone generate not The vast majority only more employment, but also more of industrial studies undertaken elsewhere have output per unit of scarce also reported that capital than do their larger counterparts. These small firms tend to use relatively more labour- results, as mentioned intensive techniques previously however, are based of production than do their larger competitors.35 on the assumption that capital is the The relative degree only scarce resource. Consequently, it is now of labour intensity of Sierra Leone's small-scale necessary to determine industrial establishments whether the industries or processes with would however appear to be somewhat apparent advantages in effi greater ciency also generate the highest return than elsewhere. Studies from such diverse countries to the proprietor or 'econ as India, omic' profit rate. These measures the Philippines and reflect the scarcity of all resources. Ghana indicate that the differences in the capital stock per worker between large- greater than 3 and small-scale industries are no ten times. , Thus, even accounting for differing defini- 3.2.4. Returns to proprietor tions, the seventeen and 'economic' profit rate large- times differential in labour intensity between and small-sce industries in Sierra Leone is significantly higher The return to the than those differentials reported elsewhere. small-scale industrial proprietors and the economic profit rate within the various small-scale industries The relative labour intensity of Sierra Leone's w:ill be examined also becomes small-scale industry in this section. apparent when the Sierra Leone results are with those generated compared The 'return to the proprietor' from recent studies of small-scale industry parameter is important because it under- provides not only a measure of the real income earned by the 66 Small-scale Industry in Employment Generation Small-scale Industry Output and Employment proprietor but also, to the extent 6, that the proprietor is a scarce facto of production, an indication of TABLE 3.4 Sierra Leone: Annual the relative economic viability mean return to proprietorand various industries and of th profit rate economic processes. The economic profit by majorprocess and by majorsmall-scale an additional rate provide Industrial industry category measure of the relative economic Annual returns viability of the vari Economic profit ous categories small-scale industries in Sierra Leone. to proprietorss Before proceeding to examine economic profit the returns to the proprietor (leones) total rate, it will first be necessary and th( capitalstock to specify more precisel] how these measures were obtained. Since one of the primary objec. Taidrigio23 tives of the study was to examine 'Trn' 982 the economic viability of small-scaic 46 industries, measures of Modern the economic rather than the financial 982 were required. Consequently, return! GTraditional' 46 it was important that 1 572 all inputs be *Modern' 120 valued at. their opportunity rather 1463 'returns to the proprietor' than the actual costs. Thus, the 100 measure for each firm was obtained Carpenters subtracting from the by Traditional' firm's value added figure the opportunity 2062 cost of 'Modern' 169 its annual capital services 2 060 and its annual non-family labour 21 The capital services figure was services. Blacksmiths obtained from Table 2.6 where the annual user or rental cost of capital 'Traditional' had been estimated at a discount 745 75 rate or 20 'Modern' 4 per cent, the assumed opportunity -850 -25 Leone. The estimate of cost of capital in Sierra the opportunity cost of the proprietor Baking family labour was and 'Traditional' obtained by combining the assumed 1 476 opportunity 'Modem' 26 costs of hired and apprentice labour. The opportunity 10601 hired labour was assumed cost of the 'Includesboth randomly and purposively sampled 25 to be equal to the firs in these -ocalities actual money wage paid, since an earlier productic' with 2000 or more 1 function analysis had indicated that inhabitants. marginal product and ti,. wage the with 20rme iata. rate of hired labour were quite similar. The production function SOURCE Survey data. the m~aginal analysis, on the other hand, product of apprentices was higher indicated that remu.cTZftion than the nominal for they received: thus, in this study, comparative purposes, the 'economic' the apprentices' expressed profits figures have been labour in rate terms as a percentage as valued in terms of the estimated value of the total value of the firm's of their marginal capital stock. TheTheuc 'economic'e omc i 42 The returns to the proprietor are summarised in ' prdustprofit r rateate measure,s on the other Table 3.4. T signed to provide ro.h th rhhand, was de- table an indication of the return n ,w sd -b reveals the wide variation in industrial generated by the by p oesad the returns to the proprietor both firm when all inputs, including process and byb industry.n utr The returns, the proprietor's input, had .T ert rs foro example, vary from a been valued at their opportunity high of Le.10 601 per year e a p e ayfo cost. On this basis, the 'marginal', for 'modern' bakers to a low yet viable, firm would thus Le.850 per year of negative generate zero returns or a zero for 'modern' blacksmiths. At profit, economic level, bakers the major industry while a firm generating a positive generate the highest proprietor's pure surplus return would be earning a return, while tailors or positive 'economic' profit.43 The and blacksmiths generate the figure for each 'economic' profit lowest return. firm was obtained by subtracting One of the most striking results presented the from the returns to in the table, however, is proprietor, the opportunity cost the indication that often the of the proprietor's labour. For proprietor's return varies markedly lack of a better indicator of this between processes parameter, the opportunity cost in the same industry. Indeed, analysis the proprietor's labour of procedures applied of variance was measured in terms of the to these data revealed that except value of the proprietor's approximate and carpentry, for gara dyeing marginal product in each industry." these variations were statistically 45 Finally, be significant. It should noted that more modern technique in each industry did not 68 Small-scale Industry in Employment Generation Small-scale Industry Outat and Employment 69 TABLE 3.5 Sierra Leone: Annual returnsto proprietorby majorindustry these activities. The estimated 'economic' profit for these firms, and locality for 'traditional'processesonly (leonesperYear) expressed as a percentage of their capital stock, has consequently Industrial Localities been estimated and summarised in Table 3.4, where the data have categories' been arrayed by major process and major industry. The most striking Villages 2 (X- 20 000- Afore tian result is that, with the exception of 'modern' less than 20000 blacksmithing, all the 250 000 250 000 processes and industries 2000 generated a positive 'economic' profit; thus, if the assumptions underlying the analysis have been correctly speci TailoringGailring 1316- 4 1146545 36365 tied, all the major types of small industrial activities, except 'modern' Gara dyeing 33 1 572 5 blacksmithing, must be considered economically viable. Carpentry 33 1315 2980 1 514 The 'economic' profit rate varied widely, however, Blacksmithing 199 77) both by proc,,zs 645 - and industry. Gara dyeing and Baking - 1443 1 carpentry, for example, generated the 564 - highest rates 'Includes both randomly and of economic profit, exceeding 100 per cent in several purposively selected 'traditional' tirms. instances, while baking and blacksmithing generated economic profit SOURCE Survey data. rates that were significantly lower.4 7 Moreover, the 'economic' profit rates differed significantly by major processes within each industry.4 8 It was the more 'traditional' necessarily yield the highest return to the proprietor. rather than the more 'modern' processes Although in that generally produced significantly tailoring and baking the more 'modern' techniques higher 'economic' profit rates in produced the these major industries. The analysis highest returns, in carpentry and thus indicates that the 'tradi blacksmithing this was not so. tional" processes The used by small-scale industries in Sierra Leone are returns to the proprietor also varied by location. This fact demonstrated is economically viable at the present time. in Table 3.5 where the returns to those proprietors Finally, it would be of interest to ascertain if using the more 'traditional' processes are arranged the economic profit by locality size. rates were related to the underlying The mean proprietor returns are generally factor intensities of the proizesses highest in the intermedi- utilised. Indeed, assuming ate size localities and lowest in that labour is relatively abundant"and the villages. For example, the tailor- capital is ing proprietors engaging relatively scarce, one would hypothesise that the more in ordinary sewing in those localities with labour-intensive processes would generate the highest r,";e of econ from 20 0(10 to 250 W0 inhabitants generate Le.465, while those tailoring an annual return of omic profit. If one compares the proprietors in the villages generate only Table labour-capital ratios presented in Le-164 per year. The relatively low return inthe villages isundoubt- 3.3 with the economic profit rates in Table 3.4 one discovers that in every instance, the more labour-intensive process generated edly due to the fact that the proprietors also engage in farming the highest 49 economic profit rate. Thus, the 'economic' profit rate of activity. Unfortunately, it has not vet been possible to determine income generated by the the the various small-scale industrial activities proprietor's farming activity, to be in Sierra Leone was seen The returns directly related to the relative labour intensity of the industrial to the proprietor for firms in localities with 2000 or processes utilised. As a result, labour intensity or capital more inhabitants, however, do shed some light on the productivity relative in- measures would appear to be useful come position of the proprietors. substitutes when profit figures The mean annual return to the are not available. proprietor from the randomly selected firms, for Le.672. The unskilled urban worker receiving the example, was The conventional factor-intensity earn approximately-hed Le.250 per minimum wage ' a partial clue measure, however, provides only year, while the farm proprietor small-scale as to the relative economic profitability of various received a mean annual income industrial firms. It does not fully capture. the role played in'1974-5 of approximately Le.475. by another potentially scarce resource, the proprietor or entrepre The mean return of small-scale than that earned by individuals industry proprietors was thus higher neur. Thus, it is important in these two roups. delineate to ascertain whether it is possible to certain entrepreneurial characteristics that may have some An examination of 'economic profit' rates of the small-scale indus- influence on the economic profit rate of small-scale industrial tries can provide additional insights firms in into the economic viability of Sierra Leone. 70 Small-scale Industry in Employment Generation Small-scale Industry Output and Employent 71 3.3 THE ROLE OF THE ENTREPRENEUR entrepreneur with access to a large amount of initial capital, which would enable the entrepreneur to Entrepreneurship is a difficult start on a larger scale and better concept to capture and specify quanti- exploit the market opportunities present. A fifth and related hypoth tatively. Although definitions of entrepreneurship abound, a corn- esis is that entrepreneurs expanding with onl, internally generated mon theme is that the 5 entrepreneur is a key decision maker. " The funds would have lower economic profits than those entrepreneurs larger the supply of such decision makers, other things being equal, expanding with the benefit of the outside capital market. The argu the better will a country's other scarce resources be combined for ment is similar to that relating to the entrepreneur's initial capital. A productive purpose and consequently the larger will be its output. final hypothesis is that entrepreneurs whose fathers were not farmers The supply of effective entrepreneurial talent, however, is limited would have higher economic profits than those entrepreneurs whose and not all firms are equally successful. fathers were farmers. The logic underlying this hypothesis is that the Thus, it is necessary to identify those characteristics of the entre- father's occupation provides an indication of entrepreneurs' social preneur that may influence his or her performance and to ascertain and psychological attitude towards industrial activity. It is corre which entrepreneurial characteristics, if any, are statistically associ- spondingly assumed that entrepreneurs will be more favourably dis ated with successful or economically profitable firms. Such an posed toward industrial pursuits and be more successful if their analysis would provide not only an indication of the potential con- fathers have been engaged in industrial or related activities rather straints to an expansion of the supply of entrepreneurs, but also than in farming. provide insights into how policies might be formulated for overccon- These six hypotheses must now be investigated empirically. If one ing these constraints and for expanding the quantity of effective assumes that these characteristics are independent of one another entrepreneurial services in small-scale industry in Sierra Leone. and influence econcmic profits - the dependent variable - in an There are several entrepreneural characteristics that can be hypoth- additive manner, these hypotheses can be investigated together by esised to have an important effect on the economic performance of statistically estimating a single equation of the following form:5' small-scale firms. Firstly, one might hypothesise that the entrepre neur's acquisition of some formal education would be expected to have a positive effect on the profits of a firm. As noted earlier, the Pr = a + b1Ed + b2Ex + b3Bk + b41C + b5RP + b6FO + si education would be assumed to enhance the entrepreneur's mana gerial organisational and technical skills and consequently influence where Pr is the return to the entrepreneur, a is a constant, Ed is a his or her ability to operate the firm. A second hypothesis would be dummy variable which is equal to one if the entrepreneur possesses that entrepreneurs with greater experience or greater numbers of any formal education, Ex is the age of the business, Bk is a dummy years operating that firm would be expected to earn higher economic variable which is equal to one if the entrepreneur keeps even ru profits than those with fewer years' experience. The argument is dimentary books or accounts, IC is the amount of initial capital of the similar to that used for formal education. A third hypothesis would firm at the time it was established, RP is a dummy variable which is be that those entrepreneurs keeping records or business accounts equal to one if the firm used only reinvested profits to finance its should be expected to earn higher economic profits than those who expansion and zero if it obtained funds outside the firm, and FO do not. The underlying assumption is a is that record keeping should dummv variable which is equal to one if the entrepreneur's father was enhance the managerial ability of the entrepreneur and thus affect not a farmer. Although the linear form of the equation has no the firm's economic more profits. A fourth hypothesis is that entrepreneurs claim to validity than any other, it did provide a useful with and conve a larger amount of initial capital would earn higher economic nient point from which to begin the analysis. profits than those with smaller amounts of such capital. The underly- The result of the regression analysis ofthe equation based on a sample ing logic for the hypothesis is that, given the imperfect capital of those 70 small-scale industrial firms that possessed the required profit market, one might expect a considerable advantage to accrue to the and entrepreneurial characteristics data was as follows: 72 Small-scale industry in Employment Generation Small-scale Industry Output Pr and Employment = -1,057.0 593.9 Ed + 56.7 73 Ex + 3,156.9 Bk + 0.37 IC (1,240) (467) training on how to (23.5) (866.4) (065) keep and use financ,i ' records +1,147.7 RP management 1 as a tool of 554.3 FO may be one effective metod the economic of assisting and increasing (542.0) viability of small-scale (559.1) industrial enterprises in Sierra The standard Leone. pThestheaderrorsfor the individualcoefficients parentheses; e rrostd for the vidua are listed in the the adjusted R?2 for ins are9 llarger The results of the analysis the equation is 0.59. amounts of initial also reveal that those firms with successful than those capital were not necessarily access to commencing business with any more Although smaller amounts. For a cross-section the 'initial analysis, the results indicate capital' coefficient provided a reasonably good tat the equation has was positive, characteristics estimate of the underlying entrepreneurial statistically significant. it was not that affect the economic returns simple technology The generally low initial capital individual to the entrepreneur. The and the general lack of requirements, characteristics must now be production involved in economies of large-scale examined in turn. small-scale industry make it to enter on a rather possible for a firm The relationship between formal small-scale and compete quite success is surprisingly weak. education and entrepreneurial successfully with Although not statistically significant, the larger firms. educational coefficient is negative, Moreover, the lack hinting perhaps that formal of access to external sources cation and entrepreneurial edu- expansion does of capital for performance may be inversely not appear to have had Similar results have related, economic an adverse effect on the been reported for the Nigerian profitability of the firms and for bread industry in the sample. The 'reinvested a cross-section of Nigerian industries." profits' coefficient was positive various There are, however, and significant at the 5 per cent explanations for this weak indicating that level, relationship between entrepre- the entrepreneurs that expanded neurial success and formal profits were generally by using reinvested education. Nonformal education even more sucessful than merely a substitute is not with access those entrepreneurs for formal education but may, to external sources of capital. provide better training in some cases, Although these re.sults'are for the entrepreneur. In certainly not conclusive,55 they addition, the more do cast some doubt on the codtention formally educated entrepreneurs that lack of capital is the principal business may undertake several barrier to the success of small-scale activities; thus their effectiveness different industry diminished.54 in any one may be in Sierra Leone. Finally, formal education The final result of the entrepreneurial inversely and basic ability may be analysis reveals that correlated in small-scale industry. was no apparent relationship there good This would occur if the between the occupation students with formal education neur's father and of the entrepre were generally offered the profitability of the establishment. nent jobs in government perma- ficient measuring The coef and large-scale industry, leaving, the dummy variable for the non-farming points out, 'only the as Harris of occupation bottom of the class to enter the father was negative and careers, entrepreneurial not statistically significant. Thus, while the bright and energetic family background - at least the individuals without formal in the formulation used in education turn to business did not seem to have this study - as the best available alternative', a major impact on the economic The years of experience small-scale profitability of of the entrepreneur, on industrial establishments appear to have an the other hand, do in Sierra Leone. important bearing on entrepreneurial This initial formulation of 'years success. The the relationship between the in business' coefficient is positive neurial characteristics and entrepre cent and significant at the 5 per the economic profitability level. The 'education' gained has made it possible of these firms while operating the firm to isolate those entrepreneurial thus appear to be a more would that seem characteristics important determinant of to be associated with a high success than the education entrepreneurial Specifically, level of economic profits. gained in a more formal years of experience and Entrepreneurs setting. characteristics the keeping of records are who keep even some that appear to be two or accounts rudimentary form of records positively associated with economic appear to be more successful profits. On the other do than their counterparts who hand, formal education, not. The 'record-keeping' initial capital level father's occupation, coefficient is not only positive and access to the capital market significant at the I per but strongly associated do not seem to be cent level. This result indicates with the level of economic that technical not profits and thus may be serious barriers to an expansion of entrepreneurship. These 74 Small-scale Industry in Employment Generation results, however, are dependent on the correctness of the underlying formulation of the analysis. Further analyses, articularly on a more disaggregated basis by locality and industry, are required before more definitive statements can be made. These results may however provide some initial guide to the kinds of policies that might applied to encourage be an expansion in both the supply of entrepre- 4 Growth neurs Aspects of and small-scale industries. Small-scale Industry An important, yet frequently missing, component of small-scale industry studies relates to the growth aspects of these activities. Nearly all existing studies have of necessity been undertaken at a single point in time, and thus have approached the issues relating to growth and change from a limited viewpoint. Recently, there have been a few attempts to assemble aggregate time-series data on small enterprises from official sources.' However, the coverage of such data is not detailed enough to make it possible to focus effectively on growth issues at the individual firm level. Yet growth information is crucial to those charged with formulating and executing small-scale industry programmes and policies. As we already pointed qut In Chapter 1 we were able to follow a panel of our 1974 industrial firms over time and thus are able to provide some insights into the growth and change in small-scale industry activity in Sierra Leone over six-year period. a Estimates of the aggregate growth of small-scale industry in size and numbers are reported on first, followed by an example of changes in their sectoral employment pattern. Next, we examine entry and closure rates, followed by analyses of the patterns of innovation and reinvestment. 4.1 AGGREGATE ENTERPRISES AND EMPLOYMENT GROWTH (1974-80) Employment in small-scale industry increased in Sierra Leone during the six-year period from 1974 to 1980. As Table 4.1 shows, however, the magnitude of this increase varied widely by locality size. In the larger urban localities, which would include Freetown and localities with over 20 000 other inhabitants, the annual rate of increase of employment in manufacturing was approximately 6 per cent. In smaller localities with 2000 to 20 000 inhabitants, on the other hand, 75 TABLE 4.1 Employment by branch annual of activity as percentage growth rates in employment of total small-scale manufacturing, and nwmber of establishments 1980, and compound in small-scale manufacturing Activity by size of locality, 1974-80 Over 250 000 inhabitants 20 000-250 000 inhabitants 2000-20 000 (Greater Freetown area)' inhabitants % (Makeni, Bo, of total Growth rage 1974-80 Kenema) (Portloko, % of total Growth Mattru 1980 rate 1974--80 % of 1980 total Growth rate 197- Ertablishnents Employmen 1 180 t nts Emplo1980ent .1980...... Food 2.0 Bakery 2.0 -4.9 21.0 13.0 0.0 32.0 49.0 33.0 5.0 23.0 21.0 6.7 Other 14.0 10.0 14.0 88.0 4 Wearing apparel . 64.0 11.0 2.4 5.0 43.0 15 .0 3.03.4 Tailoring 60.0 -0.6 1.8 2.0 4.9 42.0 -4.9 38.0 -0.3 -0.7 3.8 37.0 ShoemakingTi-dying -4.1 nO 3.0 9.5 3.0 6.2 2.0 -7.513.0 -8.5 Other 0.0 16.0 2.03.0 -7.2 -3.0 0.0 0.0 -8.8-3 Woodwork 11.0 9 ° 4.4 7.6 2 Carpentry 20.0 6.4 8.0 1.1 8.8 24.0 Carving 3.6 19.0 -3.7 3.0 8.7 8.6 0.5 15.0 24.0 22.0 -4.0 Other * -0.6 0.0 0.0 0.0 0.0 * 0.0 0.0 Metalwork 4.0 2.0 9.4 10.0 Blacksmithing 4.0 4.6 1.0 13.0 22.0 9.4 4.0 Goldsmithing 10.9 -1.9 -5.8 , 1.8 3.0 -5.6 Welding/metal 2.0 9 1 3 -5.5 7.0 6.6 Other 0.0 2.0 7.8 0.0 0.0 23.0 * 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Repairs 18.0 8.1 15.0 17.0 14.0 5.2 9.0 14.0 15.0 Watch 6.0 7.0 Radio * 13.0 9 111, Motor Vehicle 11.0 2.0 11.0 14.0 20.0 19.0 2.0 Other 13.0 17.0 17.0 22.0 2.4 7.0 20.0 17.0 0.0 0.0 0.0 manufacturing • 2.0 Total 2.8 7.2 100.0 ° 9 3.7 5.7 100.0 4.3 6.0 100.0 2.8 2.4 * Greater than zero but less than I per cent in 1950. Value for 1974 zero. 'Freetown Central Constituency umed as a sample. SOURCE Survey data. 78 Small-scale Industry in Employment Generation Growth Aspects of Small-scae Industry 79 the annual increase was 2.4 per cent, i.e. less than half as much. The TABLE 4.2 increase was Percentagesoffirms of varioussizes classified by size of apparently even smaller in villages with populations of fewer than 2000 inhabitants. locality, 1974 and 1980 " Unfortunately, the large variations from Firmsize village to village Locality size (poRulation) and the small size of the sample make it impossible GreaterFreetown area to indicate a precise employmentlocliie.2(over growth rate for these smallest rural 250 localities.2 GraerFeo000)' area 20 000-250 000 2000-20000 1974 1980 1974 1980 The annual growth in the number 1974 1980 however, of industrial establishments, did not keep pace with employment growth. In the larger urban localities (with over 20 000 inhabitants), One worker 30.6 26.7 33.0 21.8 42.0 for example, the 6-9 workers 13.0 15.6 24.9 number of small establishments 6.0 13.5 4.0 6.2 grew at a rate of approximately 4 per 10 or more cent per year; in rural 3.4 3.9 2.0 4.3 1.0 2.0 localities with fewer than 20 000 inhabitants, it Total m ay actua l ly have dec 3 100.0 100.0 100.0 100.0 100.0 l i ne d. Tot al _1 00 .0_1 00 .0_1 0 0 .0_10 0 .0_1 0 0 .0_100.0 10 0 .0 Two important questions emerge from these initial results of the 'Freetown Central Constituency used as a sample. survey data. First, what could account for the rates of increase in SOURCE Survey data. employment, numbers of establishments and firm size between and 1980? Second, what could account 1974 for the differential rates of changes in the relative increase in localities of different sizes? importance of the various enterprise size classifications are also noteworthy. The increases in employment and number Table 4.2 shows that first, the of firms in the small- percentage share of one-person scale industrial subsector of Sierra Leone cannot enterprises declined drastically in all be explained by localities over the six-year period. government policies and promotional efforts. This may reflect the lack of From the fiscal and competitiveness of such enterprises monetary point of view, government industrial within the sector. Poor business policies still favour motivation, lack of management the urban large-scale sector.4 Nor could the over-all and other skills and very lom turn-. performance of over are possible factors contributing the economy be considered a major positive 5 to this state of affairs. Second, factor. From 1974 to the percentage shares of 1980, the Sierra Leone economy as a whole the namber of small-scale firms cmploying stagnated with the GNP two to five, six to nine and increasing at a real annual rate of only 0.4 per cent. ten or more persons increased in localities Indeed, given the of all sizes. However, it is important rapid decline of employment in mining and the to note that the bulk of small sluggish employment enterprises are found in the growth trends in large-scale manufacturing, the small-scale second category of size groups - firms sector, employing two to five workers. The lowest which has little or no entry barriers, may share observed for this have become the main group was 52.8 per cent bulwark against open unemployment, at least for Freetown in 1974, while the highest share in urban areas. observed was 66.9 per cent What could account for the differential for small towns (2000 to 20 000 inhabi rates of increase in employ- tants) in 1980. Finally, ment and number of firms in localities despite their high percentage increase, firms of different sizes? It is possible employing ten that the highly uneven income distribution or more workers continue to account for the smallest in Sierra Leone, which share of the total number. favours the urban centres, creates In this size group, problems of manage substantial purchasing power in ment and co-ordination large urban locations of production activities may still constitute an and thus encourages the establishment of obstacle different to stability and continued growth. types of small enterprises there. It has been estimated6 that urban areas, which support tion, only about one-sixth of the total popula accounted for about 40 per cent of total income in 1975-6. 4.2 SECTORAL EMPLOYMENT GROWTH Moreover, our studies of expenditure PATTERNS patterns indicate that income (1974-80) elasticity of demand for small-industry quently, products is high. Conse increases in income would generate more demand for them. 7 Although our results indicate that the average As Table 4.1 shows, there were important variations in the size of small-scale ment growth employ manufacturing enterprises rates of the individual industries. In localities was increasing in localities of all sizes,8 the sizes, of all activities in which employment increased the most rapidly were 80 Small-scale Industry in Employment Generation Growth Aspects of Smal-Scale InduSry 81 the food industries. There was particularly rapid growth in baking, followed by rice, ing countries are, at best, palm oil and coffee processing. The scarce. In an attempt to shed light same trend was area, such information on this observable in repair activity was collected duriig the second Sierra woodworking. particularly motor vehicle repair Leone Wearing apparel -- especially weaving - and survey. and shoe re- To ascertain pairing - was generally the the closure rate, a sample tf 60 least rapidly growing activity in all small enterprises in 3 localities, although growth localities enumerated in the 1974 was positive in the larger ones. The survey were revisited by enumer performance of the other ators in 1980 to see if they were still branches varied widely by size of locality. operating at the same locations. Metal working, particularly In Freetown, 57 per cent of welding, was increasing .apidly our 30 sample enterprises had vanished larger localities, in the over the six-year but was declining in the smaller period while, in the two smaller localities One ones. 20 000 (2000 has to be cautious, hijwever, in interpreting inhabitants), a somewhat larger proportion, growth the extraordinary disappeared. 66 per cent, had rate observed in the bakery industry. On an annual basis, Freetown's closure During the 'hunger cent rate wa- 9.5 per season' in Sierra Leone (March-August) while, for the other localities, the closure rice is being planted but is rate was 11.1 pe c cent. not yet being harvested. Bread These enterprises had not necessarily is therefore often regarded as permanently closed, however, substitute for rice, a since they may have moved and bread output increases throughout or switched to other lines of business. period. 9 the Indeed, data from Experience gained during a field survey the 1980 survey indicated that approximately 1974-5 in Sierra Leone in cent of 2 per revealed that, as soon as rice harvests the businesses moved to another address began, the numerous each in the same locality 'traditional' bakeries ceased to year. Anecdotal information and operate. Some changed their line of entrepreneurial histories also activity, and others left revealed that changes of occupation the neighbourhood entirely. It is possible, or activity are not uncommon. therefore, that the observed For example, in one of our survey high growth rate for rural bakeries locations, 3 out of the 30 firms merely a transitory, is studied in 1974 had changed seasonal phenomenon, their line of business by 1980. A tailor The exceptionally had started illicit diamond high growth rates in branches mining, a photographer had begun work in urban such as metal rice-milling selling areas are to be expected. The boom machines, while a traditional baker had other in housing and As we expected, taken to tiradirig. types of construction, such as ultra-modern these proprietors cited better financial hotel buildings and main gain'as the offices, in Freetown and other reason for changing their activity. In urban centres creates demand for the the same locality also, four products of the metal workshops. proprietors were reported to These thus fare somewhat better have moved to unknown addresses, urban than in rural in while one tailor was reported areas. The declining trends observed to have gone back to his village. such as weaving in activities another locality, In (included in 'other wearing apparel' where the craft of tie dyeing in Tabh 4.1)and women, was dominated by shoe repair are equally foreseeable. our survey revealed that the closure 'Country cloth', as the product of rate was highest for this handloom weavers is called particular industry. Five out in Sierra Leone, is made by a few elderly -of eight tie-dye proprietors who were weavers who use cotton thread included in our 1974 sample and dyestuffs obtained from local of 30 small-scale entrepreneurs for that sources, and is usually locality had gone out of business, purchased by a few foreigners solely mostly for lack of working capital artistic value. for its to purchase fabric and For utilitarian purposes, country cloth dyestuffs. In general, the non-financial has not been cited for reasons able to withstand the the permanent closure of firms included competition of imported fabrics, which such factors as themselves to a variety of uses. lend death, illness or retirement. Entry data on new enterprises were obtained in the Sierra Leone survey for localities of all 4.3 ENTRY AND CLOSURE RATES sizes. Using age profiles from the total samplethe two oflargest enterprises enumerated The rates of increase locality groups, thein 1980, it was determined that, in in the number of enterprises , ,ect, annual rate of entry the balance in reality, mately 11 per was approxi of new enterprises entering and old cent from 1975 to 1980. It was lower firms closing. Unfor- localities: in the smaller tunately, data on small-enterprise 9.5 per cent in those with 2000-20 000 entry and closing rates in inhabitants and 6.5 develop- per cent in the villages. 82 Small-scale Industry in Employment Generation Growth Aspects of Small-scale Industry 83 One of the more interesting findings from the data is that about 35 TABLE 4.3 Innovations introducedby small-scale enterprisesof at per cent of the 307 sample firms'0 were less than 3 years leastfive old, while the years' standing in Sierra Leone, 1975-80 number of firms in the older age cohorts progressively declined. For example, only about 4 per cent were 15 years old and only 1 per cent T.pe of innovation % of respondents 30 years old. Such information, when combined with the closure rate 1.Restyled goods and services 47.6 data, indicates a marked degree of instability among small enter- 2. Started making prises, particularly within the first two or three entirely new products 38.5 years. 3. Transformed the workshop 34.5 4. Bought new machines 22.5 4.4 PATFERNS OF INNOVATION dentsNo: hadThe the sum opportunity f the percentages exceeds 100 because each of the respon to give more than one positive response. In addition to the process of growth and attrition of small-scale SOUtcE Survey data. enterprises, the nature and extent of any innovations they had made These changes tended to be concentrated in certain types during the period between the surveys were ascertained. An attempt of indus tries, including tailoring, carpentry, tie dyeing, bakery and black was made to investigate whether Sierra Leone entrepreneurs had smithing On the other hand, in some activities such as pottery, introduced new goods, new qualities of goods or new methods of tinsmithing, goldsmithing and watch repair, and rice processing, production; transformed their workshops or purchased new kinds of virtually no alterations seemed to have taken place. machines; and whether they had ploughed their accumulated profits back into their businesses. About 42 per cent (128) of the 307 respondent firms surveyed were at least 5 years old. Their answers 4.5 PATTERNS OF REINVESTMENT are in Table 4.3. Over all, the findings summarised in Table 4.3 show that at least 48 Between 1975-6 and 1979-80, each firm reinvested1" an avrrage of per cent of small-scale entrepreneurs in Sierra Leone initiated at least Le.284.3, i.e. an annual rate of profit reinvestment of L.'57 per one kind of change between 1975 and 1980. It would thus be wrong 2 to firm. When asked to explain the nature of reinvestment, 47 per characterise existing enterprises as stagnant. What is particularly cent of our sub-sample responded that they had ploughed back their striking, however, is that product changes were more prevalent than profits for the purchase of tools and equipment, 47 per cent had used machine or equipment changes. For example, according to Table 4.3, it in working capital for their enterprises, while 13.2 per cent had about 39 per cent of our sample entrepreneurs who had been in reinvested in other forms of economically related activity of business for at least five years indicated that they had started 3 the making enterprise. It is therefore clear that the bulk of our new products. Thus, a carpenter would start making sample respon cupboards in dents ploughed back their profits for the purchase addition to tables and of tools, equip chairs, or a baker would add cake to his ment and raw materials. existing product-line. Moreover, about 48 per cent indicated that It is nevertheless important to ascertain if the observed they had further restyled their existing lines of goods and services, pattern of for reinvestment is consistent with the level of economic profitability example, by creating new fashions in clothing or introducing new among small-scale enterprises in Sierra Leone. For this purpose, kinds of packaging for bread. In addition, about 35 per cent of the we shall further examine the tailoring and carpentry industries where entrepreneurs had transformed their workshops, mostly by at remod- least ten respondents gave information. In these two industries, elling the existing structures. Finally, only about 23 per cent had the nominal mean values reinvested during 1975-80 were Le.150.7 and bought new machines. These entrepreneurs cited other producers as Le. 125.4 per firm respectively which in turn give nominal annual the most important source of information on new machinery: only 10 rates of reinvestment of approximately Le.30 and Le.25 respectively. per cent of them had read about the machines before buying them. If we can assume that the mean values of capital stock in these two 84 Small-scale Industry in Employment Generation industries have remained constant at Le.526 and Le.497 respectively since 1975, then the nominal annual rates of reinvestment percentage as a of mean capital stock become 5.2 per cent for tailoring and 4.5 per cent for carpentry. In real terms, these rates become and 2.3 per cent 2.7 respectively 4 In 1974-5, economic rates on of return capital stock for tailoring and carpentry turn out to be at least 23 Part 11 per cent and 21 per cent respectively. 5 Empirical studies of the propensity of small enterprises to save Som e Empirical Case have revealed that the proportion of total profits which they or willing to save could be, of the order of one-half to two-thirds.'are able 6 If studies it can be assumed that reinvested savings, then profits serve as a proxy for business based on 1974-5 economic, profit rates - our sample tailors and carpenters reinvested only 11.74 per cent and 10.95 cent respectively per of their accumulated profits (in real terms), figures markedly below those reported for other countries. Although data the disposal on of profits were not collected, it is possible that some the extra profits of not reinvested could have been channelled into private consumption, the education of family members or other ventures. 5 The Clothing Industry The clothing industry in Sierra Leone has been chosen for analysis in this chapter because of four major considerations. First, the industry is the leading small-scale industrial subsector, accounting cent of employment, for 31 per 33 per cent of total establishments, and 37 per cent of the value-added in the Sierra Leone small-scale manufactur ing sector.' Second, the vast majority of these enterprises - 88 per cent are located ia rural areas and are thus important any elements in rural development strategy. Third, the clothing industry Leone in Sierra has both small-scale and large-scale units that compete one another with in both product and factor markets. Finally,- developing countries most have launched their import-substitution idldus trialisation policies by promoting industries such as brewing, ciga rettes and clothing2 often supported by distorted factor prices such unduly low rates as of interest, minimum wages, and overvalued foreign exchange rates.3 Such policies have in the recent past resulted in the use of capital-intensive production methods unemployment with consequences of in urban areas. The objective of this study is, there fore, to examine the efficiency and profitability of different tion processes in produc the Sierra Leone clothing industry. Major aspects of this case study include: (1) a description of some important features of the Sierra Leone clothing industry as it existed during (2) a synthesis the survey period; of five representative firms and identification of the major techniques of production to enable the analysis of the economic efficiencies of representative clothing addition firm types, in to a profit analysis of those firms, based on representa tive budgets; and (3) a summary and an examination of the implications of analytical results for policy formulation and project design. 87 88 Some Empirical Case-studies The Clothing Industry 89 5.1 AN OVERVIEW OF THE CLOTHING INDUSTRY While the majority of small-scale firms rely mostly on customers who place orders for clothing goods and serviges, the large-scale firms In Sierra Leone, clothing goods and services are produced by4 ap- sew ready-made garments and marketthese firrtsthese have through differentiated moder shoptheir proximately 17 000 small-scale labour-intensive tailoring firms con- ping centres in Freetown. Thus, tributing about 98 per cent of the entire value-added in the clothing products by adopting sales promotion techniques such as product industry, and by two large-scale capital-intensive firms. The clothing labelling, packaging and advertising, wholesaling, and retailing. In goods include sewn or knitted fabric such as trousers, jackets, shirts, effect, the large-scale clothing firms cater to the tastes of the emerg school uniforms, t dresses and so forth, while ing middle and upper class, while the small-scale clothing entrepre clothing services, mostly produced by small-scale firms, include the neurs produce for some of the middle class but mostly for the lower repair of already sewn fabric. class, such as farmers, schoolchildren, students, teachers,and capitalclerical The use of treadle sewing machines is widespread among private staff, and small traders. Where both labour-intensive there are different small-scale tailoring firms in Sierra Leone as electricity is not avail- intensive clothing firms exist, one can argue that able in most rural towns and villages. In addition therere e sewing products of different qualities in the Sierra Leone clothing industry. machines that undertake simple sewing, whereas more complicated However, empirical evidence from Sierra Leone indicates that the types may embroider sewn fabrics. These differences in the type of existing price differential (given the same fabric) is not so much a tasks performed are also reflected in cost differences - Le.100 o result of superior workmanship of the large-scale capital-intensive Le.110 for simple sewing machines of the Chinese 'Butterfly' brand firms as of additional costs resulting from assembly and handling of or Japanese 'Hedjazi' and Lc.230 to Le.300 for the embroidery the final products. machines of the Italian 'Necchi' brand or the Japanese 'Zenith' or With regard to the efficiency of resource used among the Sierra 'Hijaz'. Within the embroidery process type, there is a wide range in Leone clothing firms, two important characteristics should be pointed the cost of machines - from Le.230 to Le.1,000 for the British out. First, the measurements of observed inputs and outputs do not 'Comely' brand. reflect full capacity utilisation of plants and equipment. Indeed, the In the rural areas of Sierra Leone (localities with less than 2000 observed levels of excess capacity among the clothing firms fanged people), most of the sma!l-scale tailoring firms operate in temporary from 18 per cent to 86 per cent. Reasons for such excess capacity workshops or mud buildings and pay monthly rentals of about one include inadequate demand for products and services, seasonal fluc leone. But in localities with an excess of 2000 people, 34 per cent of tuations, and lack of, or high cost of, raw materials. Second, although private tailors operate in workshops built of cement walls, 25 per cent survey data revealed that about 65 per cent of Sierra Leonean operate in verandas, 25 per cent operate in mud houses, while about clothing firms said that they joined the clothing industry to make 16 per cent utilise temporary workshops. The average workshop profits (not necessarily maximising profits), 35 per cent of those firms space is about 100 square feet (with lower and upper ranges of 20 joined the industry for other reasons such as family enterprise, square feet and 450 square feet per workshop respectively), and father's major occupation, and security. average monthly rental is about seven leones (with a range of Le. 1 to Le.50). Apart from the large number of widely dispersed private tailoring 5.2 BUDGET ANALYSIS OF REPRESENTATIVE FIRMS5 firms, two large-scale clothing firms are in operation. One of the firms is owned by a foreign company, produces about 14 brands of In order to examine the efficiency of the major production techniques knitwear, and has fixed assets of about Le.115,000. The second firm, utilised by the different clothing firms in Sierra Leone, data obtained with total fixed assets of about Le.200,000, is a joint venture of Sierra from the 132 small-scale clothing firms included in the sample have Leonean and foreign interests and produces trousers and jackets. been grouped into five representative types as shown in Table 5.1. Each of the firms operates in permanent cement buildings, which Machine-renting firms, the first type, constitute about 7 per cent of constitute about 46 per cent of the value of the total fixed assets. tailoring firms in Sierra Leone - machine rentals cost from Le.1.50 to The Clothing Industry 91 Le.2 per machine/month in Sierra Leone. There is no organised market for renting sewing machines; a few, relatively wealthy busi -. nessmen, landlords, or farmers do occasionally rent sewing ma -- - chines. The second group of representative firms operates simple . ' machines for sewing and/or repairing clothes. These firms, which constitute the bulk of small-scale clothing firms in Sierra Leone (59 per cent), do not have the facility for :adding embroidery on sewn fabric. The third and fourth groups constitute 24 per cent and 10 per cent of small-scale clothing firms in Sierra Leone respectively. Al 'Z-though each of these does straight sewing and embroidery work, the -~-~ ." latter utilises relatively more costly machines. The fifth representative .) ,... firm type is the large-scale capital-intensive firm of which only two 00 00 t: existed in Sierra Leone during the study period. UM 0 _2 " 0- 5.2.1 Technical efficiency' Ma What stands out from Table 5.1 is that only three of the five types of 11-I representative firms are even 'technically efficient'. The straight sewing plus expensive em'roidery machine representative firm (type 4) as well ,o - rj -* * as the machine renting firm (type 1) are technically inefficient, since they require both more capital and labour to produce a unit of.qut~ut 0 0 =than does the firm type that engages in straight sewing (type 3). Uncertainty relating to the availability of equipment may be a 2. primary cause of the technical inefficiency of the machine renting type firm. In the absence of organised rental markets or any form of legal contracts binding on rental transactions, machine owners often .- withdraw their equipment either at times when sewing activity is at its _ I . peak or when the renter ought to be working. Such disruptions can "- 2 EU - have an adverse effect on efficiency. " - " .The technical inefficiency of the type 4 firm (straight sewing plus ,jV.E cc expensive embroidery machines) could be attributed to lack of fore aJ= " .sight or knowledge on the part of the entrepreneur to realise ahead of time the inadequacy of the equipment.7 Since it is probable that the ~ cc - marginal value product of these machines is greater than their salvage "- W. . value, and their salvage price could be less than what they paid for CAE 'A 0 -. them, the entrepreneurs have still retained them to minimise their f I V ".2 losses. Losses could be further minimised if the entrepreneurs could . . .U " locate their establishments in market situations where the equipment . . LV-.2 .z -- could be utilised more intensively so that capital-labour ratio would ,- a-e a 0 0 25E~ E -J Z V be reduced. 92 Some Empirical Case-studies 5.2.2 Economic efficiency 8 93 Table 5.1 shows that as the productivity for the of capital declines from straight sewing type firm to 7.2 - labour 1.0 for the large-scale factory productivity increases from .50 to .77 for these same activities. M. 00 M Although low productivity S- 1 of labour could be due to inefficiency r? (such as - utilisationpoor management), it could however be a result of an abundant resource such as labour. of a fuller ,0 reveals that the technically Indeed, Table 5.1 efficient representative firm Z generate higher levels types that :(N of output per unit '. of capital - the scarce factor, also employ more labour per unit of capital. For example, both Y/K ',. and L/K ratios "i 4 Z C 00 decline from 7.2 to 1.0 and the representative 14.4 to 1.3 respectively for firni types of straight sewing factory. It and the l:irge-scale would seem therefore that the clothing firms small labour-intensive are more economically efficient capital-intensive than the relatively ones, at least in a static sense. D 000 5.2.3 Economic profit among clothing firms Table 5.2 reveals that the clothing -% firm type carrying the highest rate I" engagesof economic profits on total capital in straight sewing using cheapcost (101 per cent) is the one embroidery machines. Thisthat is 0 followed 4% e straight sewingby the second representative machines and earns 59firm type, which utilises 0, per cent economic ratesimple of profit. Both of these profitable I other firm types are small-scale. 6-6 small scale clothing The two straight firm types (i.e., be earningsewing plus expensive embroidery machines)machine renting negative economic profits, were foundand to ,,,< which was not a surprising C4 result in light of their previously ", production. reported technically inefficiency CD in ,< ( Table 5.2 also CO highlights the lack of economic large-scale capital-intensive viability of the profit rate firm type, which of negative 27 per cent. earns an economic 00 Furthermore, in order to differen- tiate its products from those of small labour-intensive firms, extra packaging, labelling it incurs 0 prices for and other handling costs '.0 the product. It is clear froni that reflect higher Table 5.2 that this firm t devotes as much as type CID purchase of material 75 per cent of its total economic costs to the C and non-material inputs. Although the firm type that employs straight sewing machines a economic profit, earns 59 per cent rate it is less profitable tha, technique 3 (straight sewingof E C- E - " 2 u ,a 94 The Clothing Industry 95 Z,.- . plus cheap embroidery machines), which is more diversified and therefore caters to the various needs of customers. The firm "-:7 employ ing the fourth technique (straight sewing plus expensive embroidery machines), though diversified, employs ver expensive machines and -- earns a negative economic profit rate of -14 per cent. In this firm type, annual capital cost as a percentage of total economic cost is the highest (31 per cent) as compared with tl~e other firms. ;. o The use of moderately priced equipment to produce a variety of clothing goods and services in the small-scale clothing industry could be the most profitable technique of production. could Machine renting be profitable only if the existing elements of uncertainty in such E techniques are removed. Finally, - duction large-scale capital-intensive pro •~ methods earn a negative economic rate of profit and there • . fore do not justify investments at current factor prices. tL 0 5.3 SUMMARY AND IMPLICATIONS OF ANALYTICAL ?5 V; "RESULTS -' U The results of our analyses reveal that both small-scale firms c- 0 relatively using expensive equipment and large-scale capital-intensive firms have been shown to be economically non-profitable. However; those S 0 0 productssmall-scale firms utilising cheap equipment and producing diversified and services have been revealed as not only efficient but also 0 , S.. :. most profitable. -_An ( (~ .. important policy implication for our purely illusive to choose the clothing industryanalytical for implementing result is that itthe is CLQ c~ ..- 0' 0 _n 0-. UU U L 00 The Gara (Tie-dye) Industry 97 6. 1.1 Evolution of gara dyeing in Sierra Leone Gara dyeing appears to have been started in Sierra Leone around the middle of the nineteenth century. 9 Gara production was probably introduced into Sierra Leone by the Madin os and Susus who mi Gara e-d e rThe 6 nTe sG from Guinea into Sierra Leone and then settled in the North- ra (Ti Province of Sierra Leone. By the 1920s, tie-dyeing had reached Industry' southern parts of Sierra Leone, e.g. Bo and Pujehun. Today, many Sierra Leoneans, numbering over 1600 of different tribal ori gins, are engaged in gara production. In Sierra Leone, there are about 360 small-scale establishments private dyeing and 7 gara dyeing co-operatives which employ in total 2 6.1.2 The use of synthetic dyestuff about 1600 people. Gara dyeing is the most labour-intensive trial 3 indus activity in Sierra Leone. Export potential and female partici- It is most probable that gara dyers in pation are Sierra Leone depended on the also important features of the gara industry. For example, use of natural indigo dyes' until the early part of the 1950s. Natural about 18 per cent of gara cloth is exported. 4 Women constitute over indigo dyes have some disadvantages. First, 80 per cent of the total number of proprietors they are limited incolour among gara dyeing variety. Gara dyers can produce firms, and female lappas with only white, blue or input ingara dyeing.labour5 contributes over 70 per cent of the labour u idyesdark involvesblue coloura lengthy combinations. process of Second, preparing the use of natural indigo This case study attempts to accomplish the dye solution. Third, four objectives, namely: since it is difficult to control the concentration of the dye liquid at all (1)~~~decieteeitn (1) describe the ehd fgr yigi ierra Leone; existing methods of gara dyeing in Si times,Because natural of theindigo limitations dyes often of the natural indigo dyestuff, gara dy:ers (2) analyse the efficiency and profitability produce uneven results. of different production 1ecaseo he tins methods in the garamethds industry;idusry;in n te gaa nthetin dyesTuda, g ers commonlySierra Leone have begun using synthetic dyes. Today, (3) identify constraints on management, used synthetic dyes in Sierra Leone are Indanthrenthe mostand production and marketing; Caledon which are (4) discuss the policy implications imported from the Federal Republic of Germany of the study. and the United Kingdom respectively. Other dyes used by gara dyers include Procion-H (ICI) and Dylon dyes, both imported from the 6.1 THE TECHNOLOGY OF 6 GARA DYEING United Kingdom.' The word 'gara' refers to the leaves of a leguminous plant (Loncho- 6.1.3 Methods used in producing patterns carpusCyanescens), a straggling shrub oi- high7 woody climber 10 to 12 feet which grows in the Northern Province of Sierra Leone. TA'., The most commonly used methods of producing young leaves and some more mature leaves of patterns in Sierra these plants are Leone include machine sewing, collected and pounded, dried outside or handstitching with a needle, dripping moulded into small balls or stamping of wax before being 8 or candle on to the fabric, bunching (marbling), dried. In that state, the ga,-a is sold to dyers for preparing tying or binding with cord, gathering, folding the indigo dye solution. To most people in Sierra or a combination of any Leone, of the methods. The key element in these gara' refers to any cloth dye with the natutal methods is the production indigo dye extracte( of dye-resist areas on from the leaves fabric. As will be seen below, all methods of of the leguminous plant or to any cloth dyed with synthetic dye. pattern production are labour-intensive. The sewing technique is very prevalent in Sierra Leone. It can be 96 98 Some EmpiricalCase-studies done The Gara either by hand or by machine. (Tie-dye) Industry Hand sewing involves 99 the use of give designa long an irregular striated effect. needledetail mayand eitherbe marked thread or raffia. The tighter the binding on the rightThe sidespacing or shape of the of the thread, charcoal or pencil. of the fabric with thrhre ilb The cloth is picked up The folding techniqueh eitlns stitches until the onto the needle in running is often combined needle is almost completely effective results, with binding to create fabric is pushed covered, and then the especially in stripes. "Ae off in tight gathers. An lengthwise, then fabric is folded in the stitches alternative method is to folded in pleats at right half ac-oss the flat material run fabric is angles to the first fold. the thread and then to pull up the then bound with cord or The to form the gathers. cloth along ends raffia the distance from tied It is imperative that large and the width of the binding the folded at the end of stitching knots are will aetermine the to ensure that the gathers The fabric is dipped into resist pattern. Machine stitching are not released, the dye, opening out the is basically limited to sewing dye penetration. The folds for more even straight lines on folded ends beyond the binding tucks of any desired colour or one may be dyed in one Stamping width thereby stitching end dipped in one colour, with candle or wax is a them into position, the other end in a second ing different very popular method of colour. patterns on fabric. The produc- Other fabric is laid out flat methods of pattern 3 surface on a raised and production, such as near a fire. The person knotting were not common twisting, coiling doing the stamping sits practice among respondents so the near this fire, duriig candle or wax can be kept the period of the survey. the fire. The in a molten state in a container wooden block which is over dipped into the carved with several designs and then stampedcandle or wax, is 6.1.4 Designs placement onto the fabric.slightly This shakenprocess aeThe of the tois repeatedremove with stamp on excess eachwax The most areas the fabric. wax produces dye-resistant popular designs that are on the fabric. Tewxpoue Leone produced by gara dyers include the 'cow-eye', 4 in Sierra Marbling or y-eitn 'blade', ' 5-record' bunching involves two produced by the and 'LP-record' - all first, the fabric is alternative methods. In handstitching or sewing laid out on a corrugated the binding methods. With the tying and gathered iron or flat cement surface techiiiques, designs such or by hand, so the fabric bow' are as 'Apollo' or 'sunburst' is bunched into a small produced. The 'siti' or and 'rain area. This cloud design is the result marbling technique. Stamping of the method is the least of wax or candle on fabric (for time-consuming and can to fabric yields the example, up to 40 yards be used on any length of can also be used in of fabric can be dyed at popularly known 'batik' fabric combination with other once). It desigrs. may have techniques where the free to be been bound 6.1.5 The dyeing bunched. A secondin method certain areas, process appearance which alsoleaving creates a makes use of a needle other marbledsections The and thread. The fabric dyeing process varies along the thread to form is pulled up depending on the exact the gathers. The size produced, and the dye pattern to be determined by the of the pattern will be solution used. The native length of stitches and takes about 7 t, indigo dye solution gathered. This how tightly the fabric 14 days to prepare while method gives the dyer is takes only the synthetic dye solution Binding more design opportunities, 10 to 15 minutes. While is a technique commonly about the native indigo dye solution can either used among the respondents. 6 months once prepared, lasts be used separately It the synthetic lasts only or as a supplement to before being exhausted. a few days The width of the thread other methods. After the dye solution or raffia band, and patterned fabric is is prepared, the its placement on the fabric immersed in the dye bath. sion is longer The duration of immer will determine the details with the native indigo wound around of the resist line. The dye than with synthetic the fabric either in binding thread is dyes. itself to form a solid mass, criss-crossed length. a lattice back on 6.1.6 The width of theresist, thread or randomly Finishing determines thespaced width background colour that alongof the the band fabric of will remain after the The finishing process dyeing process. If the depends to a great extent tying threads are slightly pattern production. on the technique of apart, the dye creeps In the case of bound between the cords or sewn designs, the to I or cord has thread to be removed. Some respondents very carefully untie the 100 Some EmpiricalCase-studies The Gara (Tie-dye) Industry 101 cord, while others cut the cord with a knife or razor blade. Some first of these factors is the weather. The rainy season, which starts respondents simply pull the material apart - thus breaking the threads around the months of April or May and continues till September or used for making the design. With stamped fabric, the wax has to be October, reduces the amount of tie dyeing that c-ould be done. Gara melted off by dipping it in hot water after the dyeing has been production is usually done outside and rQuires some amount of completed. In general, final finishing stages include soap treatment of steady sunlight to dry the dyed material. Unless sufficient material is the dyed fabric to remove excess salt; soda and unfixed dye from produced in the dry season to compensate for the shortfal in the fabric; rinsing "n cold water; addition of starch to fabric; sun drying rainy seasons, there will be a limited supply of material for the and stick ironing, which involves beating the fabric. market. Finally, the quantity of fabric produced is affected by the availability of material inputs such as fabric and dyes of specific colour types. Shortages in these areas have been reported by the 6.1.7 Quality and quantity control problems res.pondents and have affected their capacity to meet consumer demands. One important technological problem that was observed in the indus try is the ratio in which dye powder, sodium hydroxide and sodium hydrosulphite are combined. As one watches the various dyers at 6.2 PRODUCTION INPUTS work, one cannot help concluding that lots of dyestuffs go down the drain. In fact, most dyers have ceased to use spoon measures to Broadly speaking, gara production inputs include labour (mostly combine the ingredients. Rather, these ingredients are poured into adult female), capital, material and management. the dye basin straight from their containers. Also, the fabric/dye solution ratio varies from one dyer to another and depends on the skill of the dyer and the shade he/she wants to achieve. As a result, it is 6.2.1 Labour difficult to repeat patterns with a uniform colour concentration and fastness to light and washing. In addition, most dyers lack knowledge The bulk of the labour input into gara production is proprietor and about the proper storage of dye solution. Left-over dye is often poured family labour. The proprietors contribute about 53 per cent of total down the drain rather than preserved and reinforced for later use. labour iaput while male, female and child family labour constitute Another problem is the uneven distribution of the dye on the fabric about 12 per cent, 16 per cent and 2 per cent of total labour input due to improper handling during the dyeing process. Sometimes dyed respectively."' Hired labour for tying designs on to fabric contributes cloth is hung up, and this leads to undesirable dye migration due to about 15 per cent of total labour input. Apprentice labour is negli gravitational effects. gible in gara dyeing. Most gara dyers learned the craft from their Finally, the quality of dyed cloth is affected by the presence of parents before attaining the age of 15 years. seams in fabric. Some dyers findl it cheaper to purchase pieces of cloth The most labour-intensive aspect of gara cioth production is the of short lengths or narrow widths and sew them together for a dyeing proc ss. With the use of synthetic organic dyestuff, the dyeing standardised size. The presence of a seam is not appealing to custom- process alone accounts for between 65 and 80 per cent of total time ers who would normally prefer a continuous length. Another flaw allocation. When traditional indigo dye is used the dyeing process observed in dyed fabric is the presence of broken yarns and threads. accounts for about 90 per cent of total time allocated to gara pro The removal of the threads is not always as careful as it could be. If duction."5 For most gara dyers, stick ironing of fabric constitutes the fabric is pulled apart to break the sewn threads, yarns are about 5 to 18 per cent, while bunching, folding and tying of designs!6 sometimes broken. Also, the sewn threads are not always entirely 2 to 9 per cent of the entire labour input. removed from the fabric surface. Such flaws on dyed fabric affect the Traditional indigo dyeing is time consuming. Fabric is immersed in marketing of the finished product. the indigo dye contained in a 44-gallon drum for about one-and-a Two major factors affect the quantity of dyed fabric available. The half hours. 'Men the fabric is stirred, and removed from the drum 102 Some EmpiricalCase-studies The Gara (Tie-dye) Industry 103 after squeezing out the dye. It is then spread on a cement floor or linear yds of gara output will convert to approximately a maximum of grass for oxidation to take place for about 30 minutes. The process is about 937 350 sq yds of gara cloth output annually. then repeated about three times. In marked contrast, the synthetic dyeing process lasts at most one hour. 6.2.4 Management inputs 6.2.2 Capital About 13 per cent of private gara producers in Sierra Leone had any formal schooling and over 50 per cent of gara producers did not keep The most important equipment for the traditional dyers is the 44- any kind of business records.' Only three out of the five gara gallon drum usedff for preparing the native indigo dyestuff. Users of co-operative societies kept up-to-date records of their business trans synthetic dyestuff possess dye basins. Other equipment items utilised actions. In fact, only two out of the five co-operatives were regis by gara dyers include wooden sticks for ironing, rubber gloves, tered; three were being managed by local personnel while two were kettles, buckets, baskets and carved wooden stamps for stamping under the management of both expatriate (missionary) and local patterns on fabric. Since most of the gara dyeing activities in Sierra personnel. Leone are undertaken outside, either in front of or behind residential buildings, little or no building costs are involved. In general, the annual value of capital services"8 for most gara dyers is about Le.24. 6.3 MICROECONOMIC ANALYSIS OF GARA FIRMS In order to examine the pattern of resource-use and profitability of 6.2.3 Material inputs gara firms in Sierra Leone, five representative firms have been delineated (see Table 6.1). The criteria for classifying gara firms into Expenditure on fabric constitutes about 80 per cent of material input representative firm types are the technique of production,..the institu purchases for gara dyeing. The three ingredients necessary for form- tional setup, the location and the quality of inputs that enter the ing the synthetic dye solution, i.e. sodium hydroxide (caustic soda), production process. Thus, ten out of the twenty two firms that were sodium oxide or sodium hydrosulphite and the synthetic organic initially sampled and one co-operative gara firm yielded input-output dyestuff constitute about 14 per cent of total material input pur- information that was good enough for constructing representative chases. 9 For mixing the native indigo dye solution, the important firms."' The firm types A, B, C and D represent 4 per cent, 42 per ingredients include the local gara leaves, other leaves, roots and tree cent, 50 per cent and 4 per cent of private gara producers in Sierra bark, caustic soda and a few tins of ICI synthetic indigo. Kola nuts Leone (as shown in Table 6.1). must be purchased for the preparation of kola nut dye. Textile fabrics (which are all imported) available to most dyers in Sierra Leone range from shirting and poplin, costing about Le.0.45 to 6.3.1 Efficiency of representative firms Le.0.65 a yard respectively, to the expensive cotton fabrics costing Le.1.2 per yard for the 'Omega', 'Santos' and 'Satin' brands and Table 6.2 shows that with respect to the efficiency of resource use, the Le.3.6 per yard for the 'Super' brand. 20 A recent innovation in gara firm types C and E are technically inefficient. While E uses more of dyeing is the use of fabrics such as lace, corduroy, velvet and both capital and labour to produce one leone value of gara output (as polyesters. shown by the reciprocals of the factor proportions in parenthesis) On the basis of the input-output information provided by ten of than either A, B, C or D, C uses more of each of the two resources our respondents,2 it was estimated that the gara industry could than A or B to produce one unit of output. From the point of view of produce a maximum of about 750 000 linear yds of fabric annually.2 productivity, the firm type A, the firm which uses traditional indigo If a maximum fabric width of 52 inches is assumed, then 750 000 dye, is the most productive in using both capital and labour. Also, the TABLE 6.1 Representativefirmtypes forgarafirmsin SierraLeone, 1974-75 Firm type' Number of Percentage of Characteristicsof representativefirms firms in establishments sample in the population A 2 4% Uses the traditional native indigo dye and medium-priced fabric. Located in both large2 and small towns. B 4 42% Uses synthetic organic dyestuff and medium-priced fabric. Located mainly in large towns. C 2 50% Uses synthetic organic dyestuff and cheap fabric. Located mainly in small rural towns. D 2 4% Use ynthetic organic dyestuff and very expensive fabric. Located mainly in cities. E 1 - Co-operative production unit, uses synthetic organic dye stuff and medium-priced fabric. Located both in large and small towns. Firm types A, B, C and D represent private small-scale producers. Members of gara co-operatives also operate their private gara establishments. Large towns will include localities with greater than 20 000 people and will include Freetown, Bo, Koidu, Kenema and Makeni. SOURCE Survey data. TABLE 6.2 Factorproportionsfor the gara representativefirm types Firm type Value added/capital Value added/labour Labour/capital (Le per hour) (Hoursper Le) A Uses traditional indigo dye and medium-priced fabric 67.75 (.015) 0.86 (1.20) 78.78 B Uses synthetic organic dyestuff and medium priced fabric 67.45 (.015) 0.83 (1.20) .1.27 C Uses synthetic organic dyestuff and cheap fabric 35.79 (.028) 0.68 (1.42) 52.63 D Uses synthetic organic dyestuff and very expensive fabric 12.88 (.078) 1.30 ( .77) 9.91 E Co-operative production unit, uses synthetic organic dyestuff and medium-priced fabric 2.15 (.465) 0.51 (1.96) 4.21 SOURCE Survey data. 106 Some Empirical Case-studies 107 firm types A and B, when compared to C, D and E, are most productive - in the use of scarce capital as well as in using the abundant . " t- ! -7 resource, labour. In terms of generating economic profits, Table 6.3 shows that the C representative firm types A, B and D (see Table 6.1) are generating - positive economic profits of Le.89.1, Le.179 and Le.1,491.6 and ... hence positive economic profit rates of 120 per cent, -= 107 per cent and -0 " - -- 117 M per cent respectively. On the other hand, while the firm (see Table type C 6.1) is generating a negative economic profit of about -q - 4 Le.168 (even though the returns to the proprietor are positive) and a Z " - negative economic profit rate of -142 per cei-t, the firm type E (see Z = ) -, Table 6.1) is generating negative economic profits (-Le.4,003), nega- " "n C tive returns to proprietors U. (-Le.999) and also negative economic rate - C of profit (-33 per cent). .t r. One returns tomajor explanation for the high economic profit rates and -'L' . proprietors of firm types A, B and D is u. - their high output- ' - " - - ' dye (column v) ratios. In these firms, each leone-worth of dye utilised in gara ,-. production generates a higher value of output than fii-. types . - - C and E. Specifically, firm type A reduces its cost by relying - use of on the - C= CC the traditional indigo dye. A 44-gallon drum containing 10 - lbs ,. of gara dye 2'-._-2qr.- I E' and costing about Le.10 could last for about four to six months before - -. the dye concentration is exhauzted. Also, since firm type W.8 A utilises traditional dyes, its products appcai more to tourists who j will pay high prices for those products. The ir, voe D repre- ' sents the r .. . - relatively more educated class of gara dyer;. With some knowledge Z. . of colour theory, they are able to combine primary colours to produce colour E. varieties that are non-existent in the "r market and are thus expensive if purchased. Second, the D type of 0" 'D q 'C a, gara dyers reduce their costs by devising < appropriate methods of r storing - used synthetic dye solutions, a practice not common among most gara dyers a C 10 high output-dye in Sierra Leone. In general, firm ty. -s B and D enjoy ratios because of their locational advantage. Being . situated 2:.,-_C in large urban towns, they are able to purchase dyestuffs from the dealers at reasonable - ...... prices. C:9 A close 3" 4 n S w 0 examination of column u of Table 6.3 reveals that, al- CC 0 though the firm type D utilises the most expensive fabric, it enjoys the highest output-fabric " ratio of Le.6.7 per linear yard. Being a > high-quality h entrepreneur, this firm type commands a high labour r ti fm t c s productivity of Le.2.5 (column t). Their products prices command higher not only in the domestic market but more especially in interna- tional markets. 6 ,- 0 Besides which, the firm type D represents the class C4 M 108 Some EmpiricalCase-studies The Gara (Tie-dye) Industry 109 of innovative dyers in Sierra Leone. They originated T-shirts, bedcovers, the dyeing of co-operative firm have window and door blinds, table covers and a feeling that they were working for someone export-oriented designs else rather than for themselves as members. 2 -2 addighn thestsuch caeofe as 'basket ofdctoof dyerflowers' which f the co-operative. Items. In addition, serve as gift TheThehigh thefimtpe E these categories of dyers have started experi- cost of production of the firm ty stems mainly from the menting gara dyeing high building costs and expensive foreign ibanagement with fabrics such as velvet, corduroy and polyes- staff as shown ters, which require in Table 6.3. More modest infrastructural special skills for achieving dye penetration and support and cheap local man agement staff could improve the profit situation retention. of this firm type. Although the firm type C receives positive returns, its economic profits and economic profit negative rates stem from the fact of its 6.4 MARKETING being a high-cost producer. OF GARA CLOTH Le.1.1 For example, it costs the type C firm (column w) to produce a leone-worth of gara compared cloth. When The marketing of Sierra with firm types A, B and D, firm C possesses Leone gara will be discussed at both the output-fabric the least domestic and international and output-dye ratios. Because this firm levels. Domestically, market outlets exist the lowest-priced type utilises in the homes of private fabric, it therefore sells cheaply, mostly producers as middlemen, in private fancy conspimers. to localised shops and in the craft shop However, being located in rural small towns, of the Ministry of Trade and Industry in dyestuffs is higher the cost of Freetown. At the international since purchases of materials are made from level, Sierra Leone gara is marketed small in African countries such as retailers who buy their goods from distant Liberia, Guinea, Mali, Zaire, Kenya and larger towns. Moreover, Zambia, in these firm types the United States and in Europe. incur high dyeing costs due to lack of some knowl edge of colour theory. As Table 6.3 reveals, the firm type E (which represents the co- 6.4.1 Domestic operative production of gara 28 marketing when resources are cloth) makes negative economic profits priced at their opportunity cost. Table 6.3 reveals that labour productivity Private gara producers in Sierra Leone in this firm type is least (Le. 1.1), the market their products in total cost average either one or a combination of producing Le.1 worth of gara cloth is highest of ways. First, they stock the finished output-dye (Le.1.3), products in their homes ratio is also lowest (Le.6.9) while output-fabric while buyers visit to make purchases. Sec low (Le.1.8) ratio is ond, they distribute their when compared to firm types A and D. finished products to relatives in large towns A major for disposal. Third, middlemen buy reason that could be responsible for the low wholesale from producers and productivity of labour retail the gara cloth through the E firm type is low morale due to lack of participa- their sons and daughters who travel tion in decision around with piles of cloth on their making by members of the co-operatives. Members heads or shoulders. Fourth, some do not take part male gara dyers spend 50 per cent in decisions relating to the purchasing of material of their time producing gara cloth inputs or disposal and the other 50 per cent travelling from village of dyed f bric. In one of these societies, good- to village selling their quality dyed fabric products. Finally, some private producers was purchased from the gara women at a flat retail their products in non-market market squares or through fancy shops or the determined price of about Le.3.50 per Ministry of Trade and were allowed to sell lappa. The women Industry craft shop the poor-quality fabric for themselves. The in Freetown. purchased lappa was Some private producers have adopted utilised for sewing dresses, handbags, etc. for price discriminatory prac sale domestically tices. For example, one of our respondents who and abroad. In the other gara society, the women dyes fabric with the simply received native indigo dye in Makeni sells fabric purchased for them, dyed the fabric gara cloth at wholesale prices of handed over and Le.2 per lappa to Makeni traders the entire dyed fabric to the expatriate but Le.3 to 4 to traders in Kono arranged personnel who the rich diamond-mining for the sewinp the dyed fabric into dresses, district of Sierra Leone. If one assumes that etc. for later the demand for gara disposal. The wome, received 15 cents per cloth in Kono is less price-elastic than in - linear yard of dyed Makeni, cloth produced. It is therefore easy to see that then such a pricing practice is bound to the members of type E of gara cloth maximise revenues in both markets. Apart from price discrimination 110 Some Empirical Case-studies The Gara (Tie-dye) Industry 111 between locations, private gara producers price-discriminate be- occurred regularly. As a result, these gara societies worked only two tween individual sets of buyers. Usually, tourists or foreigners end up and three weeks respectively during the survey period, thus operating paying higher prices per lappa than residents even after bargaining at low rates of capacity utilisation. has taken place. 6.5 POLICY IMPLICATIONS 6.4.2 International marketing6.POIYMLCA OS This study highlights some production and marketing constraints Gara cloth is marketed internationally by private producers in vari- which need policy consideration. ous ways. International visitors transacting business in Sierra Leone often purchase cloth. During our survey, two of our respondents reported sales, to traders from Liberia and Guinea. On the other 6.5.1 ProduLtion policies hand, gara producers also travel to Ghana and Liberia to sell their gara cloth. Gara producers have identified potential markets in On the production side, efficient production can be ensured in two African countries by sending samples of their design to relatives ways. First, if the gara co-operative societies are to produce ef living in Zambia, Mali and Zaire. By so doing, orders have been ficiently, they have to be organised along the lines of self-man placed and exports of gara developed. However, such transactions agement. In other words, members of gara co-operatives have to have been handicapped by the lack of adequate knowledge about participate in decisions relating to the purchasing of inputs, produc international markets. For example, during our survey, it was discov- tion and marketing of the finished products. In order to ensure an ered that one of our respondents had Le.800-worth of gara cloth improved organisation and self-management of gara co-operatives, impounded because of the lack of appropriate trade documents. there should be close contact 29 between government co-operative Exports of gara cloth into the United States and Europe also do take officials and the local managers and members of the coppetatives. place. But sometimes, in order to effect an urgent delivery, gara cloth Such a contact was completely absent in the co-operative societies of is shipped to the United States by air rather than by sea. Since the Pujehun, Kpaka and Yoni. Second, steps should be taken to improve landed price of air shipments almost doubles that of sea shipments, the output-fabric ratios, output-dye ratios, output-labour ratios of serious marketing problems have resulted. producing units and also attain some export production speci fications.3 These efficiencies can be introduced by the use of indus trial extension agents who possess technical knowledge of dyes, 6.4.3 The co-operatives and gara marketing fabric, dye storage, color theory and communication. Extension services should be organised and information disseminated in official The marketing of dyed fabric has constituted a serious production Crio and native languages with the help of visual aids. In addition, bottleneck, especially for gara co-operatives managed by local per- the industrial extension agents could be an adequate vehicle for sonnel. Two of the co-operative societies have extensively relied on developing the practice of record keeping among gara producers. the efforts of missionaries for the disoosal of gara cloth through agents in Sierra Leone and overseas. The missionaries have even introduced the use of dyed fabric for sewing school uniforms. But the 6.5.2 Marketing policies other two gara societies have relied solely on the capacity of the Ministry of Trade and Industry craft shop in Freetown to dispose of The problem of marketing should be tackled at both domestic and their products either through purchases by tourists who visit Sierra international levels. Domestically, the Government should take the Leone or shipments abroad. Unfortunately, such shipments have not lead in utilising gara cloth for decorating public buildings. The 112 Some Empirical Case-studies initiative already demonstrated by the Catholic sisters should be encouraged by the introduction of gara cloth uniforms at both pri mary and secondary school levels. The recent establishment of Export Promotion the Council in Sierra Leone to foster the international marketing of gara cloth is praiseworthy. This council should work co-operatively with the proposed industrial (gara) extension agents 7 The Bread-m aking industry for identifying high-quality dyed fabric, buying and fabric from collecting such While effortsthe provinces and shipping these to Freetown. are being made to identify markets for gara Africa cloth in The bread industry' is one and Europe, attention should also be directed of the five major small-scale industries in to the dissemi- Sierra Leone, contributing nation of information relating to to the dietary needs of most Sierra Leo accessibility to foreign markets. neans and employing over 4000 people, 85 per cent Problems relating to foreign currencies, exporting of whom live in ate trade documents agents, appropri- rural areas. and payment arrangements exist and are not known by most The bread industry in Sierra Leone is the small-scale entrepreneurs. Knowledge relating only small-scale industry in these to which the existing technology includes problems can be diffused through the proposed the most traditional as well as the industrial most modern. Thus, issues extension agents. relating to the choice of technique of The direct effort production and their implications for employment of Government and voluntary agencies to identify generation require foreign markets closer examination. Besides factor for gara cloth should not preclude individual prices, questions relating to location, initia- siLe of the market and consumer tive, by private small-scale producers or co-operatives tastes may be important in determining to market their choice of technique and own produce either locally or overseas. 3 Some therefore deserve close assessment. incentive mechanism The purpose should be devised for inducing all of this case study is therefore fourfold. production units to market their attempt First, an gara cloth. Increased will be made to describe the different technologies effort on the part of producers to market their of bread products baking in Sierra Leone. Second, some would not only lead to a fuller utilisation of existing analysis of technical efficiency, resourc- economic efficiency and es and higher profits, but it would also mean profitability of the different technological a training in self-reliance types of bakery for the co-operatives. firms will be undertaken. Third, the role of consumer taste and location in technology choice will be examined. Finally, the policy implications of the analytical results will be presented. 7.1 BAKING TECHNOLOGY IN SIERRA LEONE In this section an attempt will be made to describe the major 'pes of equipment utilised by the different bakery firms in Sierra Leoi.e during the survey period. Secondly, the i.ajor processes involved in bread making will be briefly describe". In general, the impact of existing baking technology on bread quality in Sierra Leone will be highlighted. 7.1.1 Major types of baking equipment The most important distinguishing factor among bakery firms in Sierra Leone is the type of oven utilised. During the period 1974-5 113 114 Some EmpiricalCase-studies The Bread-making Industry five different types of 115 bakery oven equipment were identified Sierra Leone. in Rotary peel oven These include the tunnel or The rotary peel oven fired by wood reel oven, travelling hearth oven, the or coal is very the multiple deck oven, the rotary much out of use in modern baking. peel oven, and the This overltype was designed to traditional peel brick or mud oven. overcome the inconvenience of hot Each of these oven types will be loading and unloading character istic of the traditional brick or mud overi (discussed next). It described in some detail. is equipped with a fire box from which heat is conducted chamber by means into the baking of water tubes, and Tunnel oven The tunnel hearth a circular hearth oven, referred to in some texts as on which the the travelling hearth dough is loaded for baking. oven, is the product of some Italian A major advantage of the rotary peel can bakery equipment and Amei- oven over the traditional firms. It is designed for large-volume brick oven is that at the end of ous operation,2 with a baking contiu- products bake the first per hour. The hearth comprisescapacity of 2500-8500 lbs of wheat loaded are also the first to be removed a motor-driven conveyor flour is rotated by means of a as the circular hearth passes in a straight which a reasonably uniform drive motor and reduction gear. In this line through a series of heat zones. bake is obtained. The only way, Loading of the found oven of this type dough is done at one end and in Sierra Leone in 1974-5 was purchased unloading at the other end of the by the proprietor This oven type oven. from Scotland over is amenable to automatic 25 years ago. trol. The tunnel oven is the most top and modem oven type inbottom bakery heat technol- con ogy and only one of its Traditionalpeel brick type was found in Sierra Leone oven The baking chamber of the in 1974-5. ovenhearth is built traditional insideof the bricksbaking or mud. Wood-firing takes Reel oven chamber until place directly on the The reel oven contains a relatively the oven has acquired chamber elevated baking stored and which houses a revolving reel on sufficient heat to complete the suspended. which baking trays are baking run. In this oven type Such ovens may be fired directly the fire has to be pulled from the with electricity or gas or baking hearth prior to the start of actual baking. Loading and indirectly with oil. Because of the unloading of this oven is done heat are greater high baking chamber, demands for with the in relation to production capacity, aid of peels. distribution and accurate The major and control of heat are difficult. disadvantages of this oven type technology, In terms of bakery are threefold. First, the reel oven could be since it is impossible to control considered modern and next to baking temperature, one ¢:nnot the tunnel oven. Like the case of ensure uniform loaf colour and the tunnel oven, only one reel oven bakers usually suffer losses through was identified in Sierra Leone incidents of burnt bread. Second, in 1974-5. the loading and unloading with peel in the midst of the the heat stream surging from the baking Multiple deck oven chamber The multiple deck oven is a modification introduce a baking conventional 3 of the due to heat loss time differential in the case of baking peel oven (discussed below). This resulting from inefficient loading runs. Thus, produced particularly type of oven is repeated and unloading and in European countries such checks during the baking process, as France and West undertake it is usually impossible to Germany. Although this oven type several baking runs. Third, uses diesel or electric firing, it is since the baking chamber is not usually considered 4 usually not properly modern technology and could clear of dust-like suspended charcoal be at best some baked particles, considered a borderline case between dough looks untidy, and occasionally modern and out-of-date tech- the sharp edges of nologies. the peel leave marks on the dough.5 The special feature of this oven type is that individual heart.is could be drawn from their respective 7.1.2 The distribution of oven for loading and unloading. Thus, oven chambers on special tracks typ~es and equipment the use of long peels for loading and unloading has been avoided. Table 7.1 shows the distribuzior 6 Secondly, it is possible to of Sierra Leone bakeries different temperatures maintain and oven type by location in the different baking chambers. being used in 1974-5. It can be seen multiple At least nine bulk of the that, while the deck ovens were being utilised traditional ovens (brick or mud) are in Sierra Leone in 1974-5. 1 located in rural towns and villages (locations with less than 20 000 population), the modern 116 Some EmpiricalCase-studies The Bread-making TABLE Industry 7.1 Sample andpopulation distribution 117 of bakery oven types by size of locality - Sierra TABLE 7.2 The distribution of bakeries Leone 1974-75 by type of equipment andsize of Oven type locality - !974-75 Locality with number ofpesons Major type of Fize.of locality Over 20 000 2000 Less than 100 Total equipment 000 100 000 20 000 2000 I S Population 20 000- 2000- P P S PS P S Less Total over 100000 Tunnel oven 20000 than 2000 1 1 - Reel oven I 1 - - - - I I Multiple deck - 1 I Tunneloven 1 - oven (Freeown) Reel oven 1 3 9 - - - - 1 - - 3 9 Multiple Multiple deck deck ovenKenema, (Bo, Makeni, oveo teny1pe6 Sefadu) 11 6 5 6------5 oven I - 1 6 Dough I - brake 2 2 Rotary peel - oven 2 Traditional I 1 ------Modern mixer 11 peel brick I I Obsolete 6 mixer - 1 17 or mud oven Slicer 1 - - 5 11 11 144 15 1 1 500 31 1 672 Cutter 3 11 6 17 NOTE: SP = PopulationSample Roller 11 ovensMud/brick 6 - - 17 - 11 144 SOURCE Survey data. 1500 SOURCE Survey data. (tunnel and reel oven) and semi-modern (multiple deck oven) baker- ies are located in urban towns. processes have been reduced It is also clear from Table 7.2 to seven or eight as shown some that following table. in the modern or semi-modern bakeries Some of the processes will be (tunnel, reel, multiple deck described to highlight oven types) some of their constraints. electrical andcombine the use of two or more automated types of bakery equipmentoven types and other dough brakes, such as mixers, Fermentation rollers, slicers and wrappers. Fermentation is the equipment Such complementary starting and pivotal point in the is not used by the large number bread-production line. Generally, of rural-based traditional there are two methods of fermen mud or brick oven type tation in bread-making. In household of bakeries. baking, the method of straight fermentation is utilised. Active dried yeast is fermented (dissolved in warm water) 7.1.3 The process of bread making within 10-15 minutes and all ingredients in Sierra Leone one fermentation process. Some are mixed in adopt the straight fermentation traditional bakeries in Sierra Leone process as indicated in the table. As Table 7.3 shows, there are about ten processes However, in most baking. In general, involved in bread Sierra Leone commercial baking, dough ingredients are mixed, method of fermentation is utilised.7 the sponge cut, weighed the douph is then lasts Usially sponge fermentation manually or mechanically, and set about four to six hours. In some in pans ?- . Next fermentation bakeries, a long overnight follows the baking, process is undertaken to slicing and wrapping of the bread. yield enough conditioning for So:-ie i, es the the gluten. 118 Some Empirical Case-studies 119 Mixing Mixing of dough fulfils two functions.8 the homogeneous First, mixing ensures distribution of components such as flour, yeast, water, salt, sugar, -.. etc. within the dough, and, second, helps in the development of the gluten (hydrated insoluble protein). As Table 7.3 > 0 shows, only the traditional ' brick oven baking firms use the hand- ='C mixing method instead c,. the modern mixing. The 0> i mixers ensures design of modern Q"_ I M.- that these machines blend, combine, -E compress, fold, -0 -0 stretch and push -. . 0 the dough ingredients in their c , - 4400 final state of develop- S < 2 0. ment'. Generally, machine mixing in Sierra Leone lasted between 10 _ I . l and 20 m in ute s. U'. Kneading ,. Kneading is a very important CI ~ - process for bakeries 0 '0' "0. that do 0. not CQ V CZ have an automated mixer. E 0 In this case, the dough has pushed together by to be hand in order to develop the necessa y gluten. With hand kneading, as is the case with the traditional bakeries, .- requires it a tremendous amount of human energy to produce required elasticity and tenacity the L Z C -4 -V ZC of the dough. iowever, in two Sierra Leone bakeries, a dough brake , was being used after mixing to E .. 0 achieve a smooth texture. .' , ~ 0 <2._- <5 < _ Cutting, scaling and shaping Most modern and semi-modern eries in Sierra Leone use automatic bak cutters to cut the dough into different sizes afer which - the dough is weighed to ensure uniformity. 5 However, in all the traditional bakeries, ,. the dough and desired weights knives are used for cutting E are determined by just feeling the . <2<< 5 B:ca . 2z - < n. dough on the palm of the bakeries that, hand. It is belieed among traditional with long experience, uniform sizes but it is can be ensured, all likely that some waste occurs through Z - the bakeries, the general rule this method. In almost . 5 of thumb for determining the E. 0 f5 E 'C.2. desired dough weight is to equate - r_ c- E a X an ounceC weight of dough Sierra to one f - Leone cent , 4J 15C. of bread for < E 0 < 5 <2:3- 6- = .c different sizes of bread. 2- < cut to the desired size, shaping After dough is rollers. is done by hand or by the use of < . Panning and final fermentation After the shaped dough t .o E E z is set in < E < < . F< a: pans, it is allowed to rise. The extent of process depends on this final fermentation flour quality and quantity; process and yeast level; mixing temperature. In most modern and in semi-modern bakeries Sierra Leone, the final fermentation process lasts for about two hours. However, in traditional " I E M E - " t bakeries, the final fermentation lasts for five to six I= E , E= hours during the daytime and up to C eight hours at night. U 0 V_ 4 U 2 Z 120 Some Empirical Case-studies The Bread-making Industry 121 An important advantage of a shorter dough-rising time in the modem inventories of material inputs and finished output held by the enter bakeries is that it allows the running of multiple shifts per day where prise at the beginning of the survey year. Equipment is composed of market conditions permit. machines (ovens, mixers, rollers, slicers, etc.), tools (peels, knives, pans, etc.) and furniture (chairs, tables, cdunters, etc.) that are being 7.2 THE ECONOMICS OF BREAD MAKING utilised by the enterprise. These assets have been initially valued at their original purchase price on the basis of information supplied by the respondents. Buildings reflect the value As in most manufacturing establishments, the major inputs utilised of both the building and for producing bread are labour, capital, raw materials and manage- land that are being utilised by the establishment. Table 7.4 presents the fixed cost and employment stock data for ment. Each of these aspects of production inputs in breadmaking will the different bakery firm types observed in Sierra Leone. It is import be discussed in further detail. ant to notice how the fixed costs per worker decline from Le.3,645 Labour for the most sophisticated tunnel oven bakery type to Le.133 for the traditional brick or mud oven type of bakery. Also important is the fact In the Sierra Leone bread industry, that fixed capital entry requirements for the modem and tradi 25 per cent of the labour force tional firms respectively range from are proprietors, 57 Le. 124,000 to Le.300 as shown in per cent are full-time hired workers and only 13 per cent are apprentices. the table. On the whole, the industry employed about For analytical purposes, 4000 people who actually our capital stock data have been con worked about 6 500 000 person-hours in 1974-5. verted into capital service flows using the capital recovery formula." It is important to note that the baking industry is the leading wage Material input employer in the small-scale industrial subsector, offering wages that compete with official and modern sector wages. Observed wages for The components which are combined for bread production include hired workers in the Sierra Leone bread industry ranged from Le.30 wheat flour, yeast, margarine, butter, vegetable oil, salt, sugar and to Le.50 per person/month. Such wages are of course paid in modern water. Flour is the most important component, constituting over 95 and semi-modern bakeries not in the traditional bakeries, per cent of the total value of purchased input in baking, and Another important characteristic of proprietor contrib labour in the Sierra utes about 50 per cent of the total value of gross output. Leone bread industry is that about 30 per cent of the non-traditional The bulk of the wheat flour utilised by the Sierra Leone bakeries is bakery proprietors in Sierra Leone are retirees of the modern sector produced at the flour mill in Frcetown. Since the flour mill was employment. These proprietors have not only transferred their sav- installed in 1967, milled flour imports have been officially banned ings from their previous employment to and the small-scale sector, they bakers have to purchase the locally milled flour. have also transferred their skills and managerial abilities that are so In terms of wheat flour costs, the price of flour in much needed in the private sector Freetown (cif) of the Sierra Leone economy. This has increased from Le. 15 per cwt in 1974 factor has also been responsible for to over Le.24 in 1980, an influencing the transfer of differ- increase of 60 per cent over the six-year period, ent types of bakery technology or an average annual into Sierra Leone. It takes a good deal increase of 10 per cent. On the of schooling, management other hand, the observed increases in and administrative capacity to embark on flour costs have not been counterbalanced modern business ventures. 9 by price increases in baked bread or decreases in dough weight. As a result, bakers do complain Capital of some reductions in their profit margins. Management input In this case study, capital has been divided into two major categories - working capitalo and fixed capital in the form of equipment and The bread industry in Sierra Leone far exceeds any other small-scale buildings. Working capital is defined to include only the value of enterprise with respect to the level of schooling and record-keeping 122 The Bread-making Industry 123 TABLE 7.5 Factor proportionsamong representativebakery firms in Sierra Leone- 1974-75 M 4 wh. capital ratio Value-added labour Valuefddedratio Labour-capitalratio (Le perhour) (Hoursper Le) Tunnel oven 2.5 (.40) 1.0 4Q QC Reel {1.00) 2.5 oven 4.3 (.23) 1.0 (1.00) Ch M:,tiple deck oven 4.3 (Freetown) 3.2 (.32) .6 (1.67) Multiple deck oven 5.3 (Bo, Kenema, Makeni, Koidu) 2.4 (.42) 1.8 ( .56) 1.4 C9 Rotary peel oven 15.0 (.06) .9(1.11) 16.7 - 4 Traditional peel brick oven 19.0 (.05) 5 (2.00) 38.0 SOURCE Survey data. Computations are based on Table 7.6. Figures in parenthesis are reciprocals of their corresponding coefficients. .Z. C5 knhabits of entrepreneurs. The percentages - CD Mneurs of small-scale entrepre who C.6 were bakers with some Sierra formal schooling and kept books in Leone were 33 per cent and 46 per cent respectively. 12 These. 13- ESfigures far proprietors exceed the weighted grand averages of 23 per cent and' per cent respectivelywho had some for formal schooling and 17 , the entire small-scale keptindustrial books. subsector In terms 8 of business " organisation, almost 50 per "' c -" cent of all the modem bakeries are private liability companies who kept systematic books i" and were subject to periodic internal auditing. It should be remembered, however, that with traditional bakeries 4) "-. dithe story is different. Most of the traditional bakery entrepreneurs did not have any formal schooling, never kept books, have any legal business status.' 3 and did not -J 0 U 0> .0 7.2.1 Micro-eonomic analysis of bakery c""4 M= firms In order to examine the pattern of resource 0"bakery 0 0 use and profitability of firms in Sierra Leone, six representative firms have been "'3"0 0 -V- - -I VU delineated (see Table = ,Z -a)0 baking firms 7.5). The criteria for classifying the bread J 0 into representative firm types are ogies of bread-baking in Sierra Leone. the observed technol g~ ~ Y 0Vac,-o 124 Some Empirical Caye-studies 125 7.2.2 Factor proportions C _.; As Table 7.5 shows, the representative traditional peel brick oven bakery firm has the highest labour-capital ratio (38) and also the C0,4, 0 t3 highest value-added-capital ratio (19). Thus, compared the to the rest of firm types, the traditional oven bakery employs more of abundant labour and generates more value-added 0 P M . per unit of scarce capital t., .0 respectively. The traditional bakery firm type is followed by the rotary peel oven - type which has a labour-capital ratio and a value- -0 Q added-capital ratio of 15.0 and 16.7 respectively. We examine below " whether ... '. - + this pattern of resource utilisation indicates the efficient performance I - - on the part of the traditional labour-intensive technology. .8 8 F5 7.2.3 Technical efficiency ' -. Table 7.5 presents, in parenthesis, the reciprocai. of the factor C.c r4 a,. - U . .. proportions shown in the first and second cc:umns. The reciprocals M show how much . capital and labour are required to produce one unit en - of value-added. The table thus 0 shows that the tunnel oven and 0. multiple deck oven are both technically inefficient. Whereas the " tunnel oven 0.- U 0 uses the same amount of labour and more capital than ,. .. the reel oven bakery to produce one unit of value-added, deck oven bakery the multiple z uses more of both capital and labour to produce E 1! . one unit of value-added. Besides these ' .obakeries, the rest are -,0 technically efficient combining varying amounts of capital and labour , - to produce one unit of = & " value-added. 2 e* "a E . '6 7.2.4 2 Economic profitability of representative bread-making firms 0 c - a Table 7.6 shows that the rotary peel oven firm type, though using obsolete technology, an ,8 represents the most profitable tLchnique of . . . . 1,.0 producing bread in Sierra Leone during the the highest economic 7 survey period, attaining profit rate of 2 2 percent. In terms of economic profitability, the ., rotary peel oven is followed by the traditional peel c0" 5 E brick or mud oven, A .-1 0 t-. , reel oven and the multiple deck oven outside the capital city, Freetown. V , > - "d --Z 2 While the tunnel oven makes a low level of economic profit of 3 per cent, ,u'u E negative rate of profit of -10the per multiple cent. deck oven in Freetown has ,,0 26 E -, CO 0 E D C; En *;E5 '. = 0 0 CI E o 126 Some EmpiricalCase-studies The Bread-makingIndustry 127 7.3 PATTERN OF TECHNOLOGY ADOPTION IN THE or readily supplying the needed spare parts. Hence, one multiple SIERRA LEONE BREAD INDUSTRY deck oven bakery in Sierra Leone closed down during A surprising aspect of our the one-year analytical results is that, although the survey period owing to equipment breakdown. obsolete rotary peel oven bakery firm type seemed advantages of the highest rate of econon.ic to combine the profitability, reasonable 7.4 LOCATION level of labour intensity FACTORS, BREAD TECHNOLOGY in production and high quality of final AND product, only one bakery QUALITY time of the survey. of this type existed in Sierra Leone at the International equipment It is significant to note that, companies in Sierra Leone have a in Sierra Leone, the modern and ency to advertise the tend- semi-modern bakeries modern or semi-modern multiple deck which utilise tunnel, reel and multiple deck type of bakery rather oven ovens dominate bread production than any other less capital-intensive and in Freetown. Freetown, the capital prestigious type less c;ty of Sierra Leone, has of bakery equipment. The dlite, on the a population of about one-quarter of a have the level one hand, million. At an intermediate of literacy to understand the procedures level of locality size, i.e. towns with international required in population range of a business transactions. 20 000 to 100 000, about 35 per cent of bakery On the other hand however, establishments are semi-modern the mass of traditional bakers who multiple deck oven type of bakeries. have little or no schooling At the level of rural towns and cannot comprehend the mode of modea villages with a population of less than about 20 000, only traditional brick technology and the required level of or mud oven type of bakeries do not have access, politically, business transactions. They also to sources of local commercial financ- abound. ing and foreign exchange. It would seem, therefore, As a result, their entry into the that in deciding what constitutes appro industry is limited bread priate technology, the to the traditional technology, a level extent of the market and the tastes of that is of operation consumers have to the probably consistent with their management be considered. Freetown has a larger .proportion capacity. and technical of government officials, expatriates and the dlite of the private sector, As well as promoting who earn higher incomes the modern or semi-modern technology than the remainder of the popillation. It bread production among of would seem that this stratum the 61itist businessmen in Sierra Leone, the of the population is able to afford the international equipment luxury of high quality bread that companies have made the adoption of such is well advertised, already sliced, technologies easier by kept fresh and moist in wax or financing the sale of their equipment on credit polythene wrappers and sold in in Sierra Leone. For supermarkets and local shops. example, one of our modern bakery respon- In Freetown, the dough weight per dents imported a Sierra Leone cent is less than set of the multiple deck oven bakery equipment the one ounce that is typical of the from France. As part traditional bakeries in Sierra of the terms of agreement, the Sierra Leone Leone. Thus, Freetown bakeries can baker was required afford to reduce dough weight to pay the sum of Le.1000 per month for in order to maintain a high quality of three years to a period product; retain their a French company while the equipment was clientele and enjoy at least a minimum rate teed for guaran- economic profit. Also, of a period of ten years. This procedure of since they produce for relatively high-income tated transaction facili- families, bread the import of foreign technology in a situation denominations range from 25-50 cents. foreign exchange of stringent Outside Freetown, control. Again, only the dlite and reasonably where the extent of the market is limited cated entrepreneur edu- bread quality of less and can have access to this pattern of consequence, traditional bakeries dominate importation. technology industry. These produce the heavier weight bread for the same amount While the equipment of money as in Freetown, companies make considerable efforts in lower denominations of 5-20 cents, and diffuse the knowledge to generate a high rate of of modern Western bread technology, economic profit. In these bakeries it may not have they probable that the percentage is made adequate provision for servicing the gain due to savings in cost as a result equipment producing a lower quality of undifferentiated product is greater than the 128 Some EmpiricalCase-studies The Bread-makingIndustr. TABLE 7.7 129 Growth rates in the number of bakery establishment employment in some and TABLE 7.8 Observed localities of Sierra Lec ie, 1974 and 1980 energy costs of baking in five bread-making enterprises - Sierra.Leone, 1980 Locality orgroup Percentagegrowth rates Firm Location Type of (approximate whole numbers) Cost per No. of onEergy source No. oven used cot pr of establishments Employment cwt( bag bags 1974-80 Annual 1974-80 Annualmot month Freetown Fretow Au 1 -36 -6 -51 -9 A Bo deckMultiple dieselElectricity Bo 475 oil and 3.1 175 79 327 55 Makeni 40 B Bo Multiple Kenema 250 7 -19 -3 Electricity and 2.6 162 42 119 20 deck Mattru 25 C Kenema Multiple Electricitydiesel oil Segbwema 4 217 36 and 2.9 42 -11 -2 -38 deck diesel Port Loko -6 D Bo Water oil 125 21 Wood 0.5 60 Bo, Makeni, 53 9 Kenema 142 tube, rotary 24 131 22 peel E Mattru Mud Mattru, Segbwema Wood 1 8 and Port Loko Wood 1 8 29.4 5 22 4 SOURCE Survey data. SOURCE 1974 and 1980 survey data. percentage loss from producing were grouped together according to heavier weight bread for the same population range, then Bo, amount of money as Freetown Makeni and Kenema (which bakeries. Such traditional fall within the population range would seem ovens 20 000-100 000 inhabitants) of to be more appropriate for utilisation and Mattru, Segbwema and in rural towns and which Port Loko fall within the 2000-20 000 range) villages. record pqsiti~e annual growth rates of firms and employment as shown in the table. On the other hand, Freetown 7.5 COMPARATIVE records negative annual growth rates of-6 and STATIC ANALYSIS OF SIERRA -9 per cent LEONE can one explainfor both number of establishments and employment. BAKERIES the observed differential growth patterns? How One important explanation for the decline in the number of bakery In this section we sh- II first attempt to examine the growth patterns firms and employment in Freetown with respect to the number with the technology being lies in the problems associated of firms and employment in the different utilised. Out of the 11 modern bakeries technological types of bread that were observed in Freetown making firms in Sierra Leone. Second, in April 1974, only 7 were found in we shall undertake a comparison April 1980. Two out of the seven of analytical findings of studies of modern bakeries had resorted to bakery firms in some other the use of a wood-firing oven as opposed African countries. It is hoped that our to using electricity, diesel analysis in this chapter will shed and gasoline as fuel. While one further light on the results of the had built a separate brick oven which Sierra Leone bread industry it now uses; the other had study. converted its electrical oven into wood firing by installing a firing box at the rear of the equipment and Sierra Leone water bakeries in 1980 Table 7.7 reveals tubes for that, apart from was distributing heat in the baking chamber. Freetown and Segbwema, accomplished by an engineer who was This conversion positive rates of growth of flown in from Lebanon at prises and employment were bakery enter- a cost of observed for the rest of the sample over Le.20 000. localities. Data for Bo revealed Table 7.8 shows the energy costs the highest annual rates of growth of of producing bread from a 100-lb firms and employment of bag of wheat flour for five firms 79 and 55 per cent respectively. If localities which were able to give reliable information on energy consumption. While these firms may not be 130 Some EmpiricalCase-studies The Bread-making Industry.: representative 131 of Sierra Leone bakeries highlights this tentative information the fact that, at least for these firms,The country semi-modern bakery costs about studies are unanious in revealing flour as the single three times the energy utilisation in the traditional brick/mud oven bakery most important ingredient in baking. In and about six times the cost in Wangwe's and Kaplinsky's energy use in a water tube, studies, flour cost constitutes rotary peel bakery oven, 75 and 76 per cent respectively of total Our 1974 analytical results cost of production. The Sierra revealed that this last type of bakery Leone sudy and Kilby's reveal that, constituted the most with respect to raw material economically profitableast tecology cost, flour alone constitutes 95 and Although the proprietor per cent respectively 90 of thisbakey ha bakery technology. of total ingredient cost in bread Althougm ther, the studies point out making. Fur in more moderne o reequipment, ipent o hihe revealedee ryal d planned in 1975 to invest the continuous upward price trends per in our recen interview with 100-lb bag of flour. Kilby recorded a 114 him that, on the between 1946 and 1961. Our Sierra Leone per cent price increase basis of his cost calculations, it was not advisable study rfveals a 60 per cent him to adopt for price increase between 1974 a technology that refies on imported energy and 1980. Each of these studies also Likewise, sources. reveals that, because the proprietor of the modern bakery of the reliance on white flour in in Kenema confessed nutritional bread making, that he was quality of bread is low in countries very little being out-priced by small traditional bakeries where incidence of on wood as an energy source. who spend malnutrition is rampant. Finally,an tii n sr m a t it is clear from the various studies that from the point 7.6 view of economic p. ofitability, the least of SIERRA LEONE BAKERIES IN THE capital-intensive brick oven AFRICAN bakeries were superior to the more technically advanced CONTEXT bakeries. seems Detailed Dissimilaritiesto conflict Evidence in respect from case studies of the bread industry in to energyNigeria, consumption. Sierra Leone and Kenya African countries are and Sierra In the Nigeria 6 Leone studies, the cost of wood firing in very few. 1" However, traditional ovens attempt will be made to onfocus the basis of available information, an attention on the important similarities per 100-lb bag of flour is much less than it is for oil-firing and dissimilarities of bread-making the contrary, the Kenya study reveals ovens. On technologies in Nigeria, Tanza- that the tunnel oven cgnsumes nia, Kenya and 7.93 shillings worth of energy Sierra Leone. The result of such comparisons per bag, while energy consumption in could brick ovens costs 9.01 shillings point the way to per bag. It is possible that the relative oping likely changes in policy and programmes the bread industry in Third World countries, for devel accountscarcities of wood vis--vis for this discrepancy. petroleum in the various countries could Similarities The results of the country An important difference exists studies reveal that a wide Sierra in the results of the Kenya range of technology options Leone studies in respect to the 17 and exist in the bread industry, the least and location of bakeries. While ovenbakrieIfKily'sstud resectvel. the result of the Sierra Leone study most sophisticated being the traditional cold erveas bech-bakery reveals that there are far more mud oven and the tunnel the esuestablishmentsl she i rin ra rural villagesv als oven andatowns, t thethe Kenyae yastudymore bakeries respectively. If Kilby's study could mark, it would seem that rapid technological serve as a bench- shows that change has taken place most bakeries in Kenya are located 8 in the African bread industry. Thus, what Kilby referred This discrepancy could be due both in urban centres.' to as a coverage and the population cut-off pointto the for extent urban ofand geographical rural areas modern oven in Nigeria in 1965 is now obsolete equipment in Sierra utilised for the respective studies. Leone, Tanzania or Kenya in 1980. However, it is clear in the country very It is very important to note that studies that urban bakeries are larger slow progress has been made in the continent in size than rural. in producing Finally, it is claimed in the Kenya study that bread quality (in brakes,bakery equipmentmixers, etc. prototypes It is such as improved brick not clear how much progress hasovens, been doughmade terms of nutrition, texture, in Nigeria where the technology being utilised.taste, neatness, etc.) does not depend on local manufacturers took some initiatives On the contrary, the Nigeria, Sierra in the Leone and Tanzania studies 1960s to produce bakery equipment prototypes.'" reveal that quality differences do exist. Kilby points out the role of dough brake in giving bread its distinctive 132 Some EmpiricalCase-studies qualities. 20 The Bread-makingindustry In terms of various attributes, 133 therefore, it would seem that bread quality differs according technique continues being used. to the type of technology to be the most appropriate though method of roducing suboptimal m t bread in rural areas of Sierra Fourth,o ofpr the d cn br a in r rl re Leone. diffusion of bread o S e ra L n . technology knowledge the use of modern and the transfer of bread technologycapital-intensive to Sierra 7.7 IMPLICATIONS OF bakeryLpne have focused more on THE BAKERY CASE STUDY equipment. Given The bakery case employment the study has at leasi five implications. potential and profitability of traditional ditional technology, which First, the tra- Fifth bakeries, they is the most labour intensive, also deser-:eand finally,promotional the production of be making an optimal may not an upgrading attention. any technology which has to be use of abundant labour and scarce of the traditional oven should the Sierra Leone bakery, capital. In ability of be based on the avail the technology which optimises local raw materials and skills. In use seems to be an outdated on resource of utilising this respect, the technology western te_2nology which burnt bricks for building bakery tween traditional lies in be- to and ovens has to be resorted and modem/semi-rodern technologies. improved upon. The major emphasis mal intermediate This opti- capacity should be to increase the technology combines the advantages for erecting different size brick bread and of high-quality wood ovens, and to improve on the considerable turnover of the or coal firing system by modern bakery with much introducing better chimney designs, adequate insulat. and less capital intensity. heat regulation methods. Second, although our economic analysis shows the outdated ro tary peel oven technology to yield the highest rate of economic profitability, it should be pointed out that the extended use technology today would face of such a a number of constraints. Such equipment may not be obsolete found even in today's used equipment kets. Even if this type mar of equipment were to be produced countries, the cost of production in advanced would be close to the current of semi-modem multiple deck prices ovens. This is because the technology of incorporating the rotary system and water tubes remains intricate and sophisticated. If on the other less developed hand such ovens were produced countries, an extra cost in scarcity element could arisL due to the of raw materials needed to importation build such iron ovens and the of the skills needed to produce pilot project prototypes even on a basis. On the whole, the production firing metal oven of such a wood may not be technically or The tunnel economically feasible. oven likewise cannot be considered nology for use, even an appropriate tech in Freetowr., owing to its large energy costs. The capacity and high result of a recent survey in Sierra that, due to the high Leone shows cost of electrical and oil energy, bakeries have already some modem reverted tn the use of brick ovens. 2 Third, both the traditional ' rural-based technology and modern multiple deck oven bakeries semi cater seem to be appropriate and to for different segments of the consumer costs of producing market. If the current a metal/iron wood-firing rotary totype would be prohibitive, peel oven pro then the traditional labour-intensive Part III Policy Assessment 8 A Review of Policy Environment, Programme and Institutional Support Several of these empirical and analytical results should prove to be of value to those charged with formulating policies and programmes with respect to small-scale industries and handicrafts. In this chapter, the overall economic policies and small industry programmes of the Sierra Leone Government will be reviewed to provide insights into the efficacy of the major policies and programmes that influence either directly or indirectly small-scale industries, and in the context of empirical evidence already presented. 8.1 ECONOMIC POLICIES The two major policies designed to directly influence industries - credit small-scale and training schemes - will be considered first, followed by an examination of several policies that indirectly affect these industries. 8. 1.1 Credit policies One of the most commonly articulated and utilised methods of directly affecting small-scale industries is through the provision capital or more of specifically credit. It is thus of importance to ascertain the need for such a policy measure and the form it might effectively take. most The findings of this study reveal that in Sierra Leone self-financing has been of overwhelming importance in this sector. Personal and family savings, for example, accounted for approxi mately 80 per cent of the funds used to establish small-scale industries in Sierra Leone, while almost 90 per cent of the funds used for 137 138 Policy Assessmen, A Review 139 expansion were reinvested profits.' The crucial question is whethe the overwhelming use of self-financing Under the Credit Guarantee reflects an underlying capita Scheme, lur example, the Bank of shortage and a corresponding Sierra Leone guaranteed loans need for improved access to credi made by.ciommercial banks and the through the formal National Development Bank financial system. The small-scale entrepreneur to small private sector establishments.4 themselves generally argue that However, the Credit Guarantee Scheni the shortage of capital and credit i does not cover single loans one of the primary constraints below Le.500 and therefore they face. Indeed, approximateli is not accessible to numerous small two-thirds of the entrepreneurs borrowers who need Le.50-200 interviewed in the Sierra to purchase items ranging from small small-scale industry Leom tools to sewing machines. study felt that the shortage of capital Although the National Development was the 'greatest or credi has loan schemes Bank difficulty encountered in 2 for 'small borrowers', such credits their business'. It i to are provided up important to ascertain whether Le.5,000. Efforts are now under way the perceived difficulty was real o to introduce a rural banking whether it was simply an easy system within the operations of response to the interviewer. the Bank of Sierra Leone. It is There is, for example, some envisaged that through this and force to the contention that capital other alternative mechanisms very and credit may small loans can be made available not be a crucial constraint for small-scale to rural artisans without cumber The initial capital requirements industry, some administrative 6 for the kinds of small-scale industries hurdles. examined in Sierra Leone are There is considerable force to the relatively modest, generally less than argument, however, that this Le.100. Moreover, the results credit to small-scale industries should of the analysis of entrepreneurial be provided at higher rates of characteristics indicated that interest than are currently firms with access to only small amounts being charged by the commercial banking of initial capital were as successful system. The present study, for example, in generating economic profits as has revealed not only that those with larger amounts of small-scale industries generate initial capital. The same analysis also a high rate of 'economic' profit but revealed that those firms that also that 'process choices' exist expanded with only reinvested profits within the major small-scale indus were apparently more successful tries. Thus, an unduly low or subsidised in generating high levels of econ- rate of interest may result in omic profits than those that more capital-intensive processes had access to outside funds. Finally, and industries being established there is evidence that excess than would be the case with interest capacity exists in all of the major rates that reflected mote closely small-scale industries examined the opportunity cost of capital in 7 in Sierra Leone. These findings, Sierra Leone. However, there is while not conclusive, pro ide the need to further ascertain under some indication that capital may not what circumstances and to what necessarily be extent changes in factor prices the overriding constraint facing small-scale or other policy variables could result industry.' in increased employment These considerations, however, or earnings. do not necessarily obviate the need for policies and programmes designed to improve the small scale industry's access to market in Sierra Leone, the formal credit market. The capital as in most other African countries, is highly 8.1.2 Policies for deliverinp 'fragmented' technical and managerial assistance the commercialwith artificially low interest rates existing sources and unduly high rates on credit from available existing on credit The results from sources in the non-commercial of this study also provide some insights sector. Since the com- policies into the efficacy of mercial banks charge a minimum designed to improve the technical interest rate of only 12 per cent for and managerial skills their most risky loans, it is not within the small-scale industry surprising that they have limited their sector of Sierra Leone. The appren lending almost exclusively to ticeship system serves as the primary their traditional customers in the trad- vehicle for providing technical ing sector and have made training in small-scale industry; very few loans to manufacturing establish- indeed, 90 per cent of the proprietors ments. The interest had previously served as rate in the market thus does not apprentices. None of the proprietors rationing instrument. serve as a received formal had There are, however, virtually vocational training' and only no formal institu- received one quarter had tional mechanisms outside the any formal education. The regression commercial banking system for mak- analysis revealed, how ing credit available to small-scale ever, that there was no correlation firms, particularly those located in between the economic profitability rural areas, of the firm and the level of formal education. Unfortunately, no information exists about the relative cost and benefits of alternative 140 Policy Assessment A Review methods 141 of providing technical training.9 In view of the importance and (3) exemption from ubiquity of the apprenticeship system, some import duty on materials kised in constructing be paid to programmes attention should perhaps the plant designed to upgrade and expand it. In addition, and on capital equipment; there (4) complete or partial exemption is some evidence to indicate that policies designed from import duties 'on raw or to provide semi-processed materials managerial training may required' in production." study has be of some value. The small-scale industry revealed, for example, that only 17 per cent of keep a the proprietors Since its introduction rudimentary set of books or accounts. Moreover, in 1960, 33 enterprises have been eligible for the regression these tax incentives, and analysis has indicated that all were firms planning to invest in excess of those a high positive correlation existed between firms that keep books and the level of 'economic' Le.50 000. policies profits. Thus, By not permitting smaller designed to improve both technical and managerial firms to qualify for similar exemptions, be of skills may however, smaller establishments importance. were indirectly piaced in a less In addition advantageous position when to these policies designed to directly influence compared with the larger establish scale small- ments. The import duty industries, there are several policies that indirectly exemption, in particular, would appear to scale industries. affect small- have created large distortions As well as those policies designed to and had a relatively large negative influence the effect on small-scale industries.'" infrastructure, large-scale industry and agriculture, The previous analysis has revealed, influence fiscal policies for example, that the small-scale industry sector- 0 except for blacksmithing, imported intermediate often in unintended ways.' inputs represent Each of these policies must be briefly examined, from 10 per cent to over 50 per cent of the small-scale industries' major output value. Since substantial duties are generally levied on these intermediate 8.1.3 Fiscal inputs, the smaller firms must policy pay a higher price for these materials and consequently higher costs than do face relatively their large-scale tax-exempt counterparts. A Fiscal policies often have negative effects on small-scale enterprises. similar argument holds with respect to These industrial establishments. the capital inputs imported by negative effects are not necessarily the The small-scale establishments are rq design result of conscious quired to pay but rather arise from the indirect effects of import duties on their capital goods while the policies designed to establishments larger achieve other objectives, such as the which obtain capital at unduly low interest rates encouragement of large-scale are already and industry or the generation of governmental often too capital intensive are often able to revenue. In Sierra Leone, duty entirely. escape this two components of fiscal policy - the Development Thus, the import duty exemption for both intermediate 1960 Ordinance of and capital goods inputs and the New Development of Industries contained in the 1960 Development Ordi Act, 1981 - together nance would appear with the import duty structure, to have an adverse, indirect effect on small-scale indirect have had a particularly deleterious effect on small-scale industry and must be _xamined industries. detail. in more The structure of import duties in Sierra Leone reinforces the The adverse impact of the Development Development Ordinance of 1960 vas a Ordinance on small-scale indus nance, tax incentive ordi- tries. Tariffs appear similar to those found widely in other African, to have been established with revenue rather Latin Asian and than resource allocation American countries, that was designed to encourage or over-all development as the primary investment industrial objective.' 3 Consequently, in Sierra Leone. The firms qualifying for these many capital and intermediate goods were incentives required by industry eligible for: are taxed at rates similar to those levied on 'luxury' consumer goods. The most frequent rate of duty levied in (1) exemption from income taxation during a tax holiday Sierra Leone is 36.5 per cent, applied period that 3-uch as toys and smoking to semi-luxury consumer goods may extend from three to ten years; pipes as well as to many capital and (2) intermediate goods deferral of depreciation allowances until after required by small-scale industry.' 4 The imported the expiration of dyes used as inputs the tax holiday; by the gara dyeing industry, the buttons, needles, thread and scissors used by the tailoring industry, and the polishes 142 Policy Assessment used by the carpentry industry, A Review for example, are taxed at the ad 143 valorem rate of 36.5 per cent. the potential large-scale competition In addition, the grey and bleached of these small firms. Those white cloth used as intermediate industries such as gara inputs by the tailoring dyeing that currently serve a relatively dyeing industry and gara market area might large is subject to an import tariff of approximately find their market positions somewhat cent."5 Finally, 22 per an improved enhanced by imported sewing machines, the basic road network. The majorkzy of small-scale the tailoring capital input of Sierra industries in industry, are taxed at 16.5 Leone, however, are primarily serving carpentry, per cent while imported only the local market. blacksmithing and baking capital Moreover, since an earlier production equipment is taxed at analysis yielded no evidence of 36.5 per cent. The relatively high economies of large-scale production tariff rates for these items when in these industries, a larger combined with duty exemption market area may not be crucial granted to most of the large-scale for them. An improved road firms creates undesirable, work, on the other net and undoubtedly unplanned, hand, may provide some added benefit effects, particularly allocative indtstries such to those with respect to small-scale as gara dyeing that require large industries, chased quantities of pur inputs. These improved roads, however, may also open or expand the markets for large-scale 8.1.4 Infrastructure policies firms. The degree to which the small-scale industries substitutable the would be affected depends importantly on how small- and large-scale products are It would appear, however, for one another. Policies designed to ellpand the that small-scale omy will also indirectly affect small-scaleinfrastructure of a developing econ- industries serving and industries. The direction tively large market areas rela magnitude of these indirect effects, ments, such as gara dyeing,and with relatively large input require however, cannot always be may tend to benefit from an improved predicted with certainty. The provision road network, while those industries for expansion of electricity, such as tailoring serving local water or roads, for example, would markeis with relatively modest appear to benefit small-scale input requirements may not. Thus, firms. These same amenities, the direction and magnitude however, also benefit the large-scale of the indirect effects of programmes firms, which may now be designed to improve the infrastructure able to enter markets previously dominated would appear to Jvary from one small-scale industry group by small-scale establishments. Indeed, to another. possessed one differential advantage by small- over large-scale industrial do not require large firms may be that they amounts of potentially costly infrastructure.16 The lack of large 8.1.5 Large-scale industrial quantities of water policies an important constraint on the likewise does existing small-scale industriesnot appear in to be Leone. The only Sierra Policies designed establishments using water were to encourage the development of dyers, but bakeries and gara try large-scale indus the amounts needed b even these also indirec:ly affect small-scale industry. industri: were :'ot small-scale The negative effects on industries stemming from the large. various fiscal incentives and These results would thus also cast some doubt on the effectiveness credit subsidies of industrial estates as viously. In this section,accorded large-scale firms have been a device for promoting small-scale however, the influence of the noted pre of the type found industries between small- direct linkages in Sierra Leone. Since the provision and large-scale industries must ture, such of infrastruc- Large- be considered. as electricity and water, is one of and small-scale industries are connected the important objectives linked through both income with the establishment of and production relationships. In industrial estates, the indication Sierra Leone, however, these link that such services may not be ages are minimal. Since large-scale critical for many small-scale firms industries are still such a makes the case for industrial tively small component rela estates less compelling. 7 of the economy in Sierra Leone, Finally, the results increased income the of the Sierra Leone study indicate generated by an expansion of improvement that the have relatively large firms would of roads might be expected to little effect on the demand on have a differential effect for the products of small the existing small-scale industries. scale industry. Moreover, the results The outcome depends import- of the study have revealed that antly on the extent of the market, the production linkages are also the relative input requirements and limited in Sierra Leone. There are virtually no forward linkages yet established from the small- to the 144 PolicyAssessment large-scale sector and A Review 145 thus the large-sce provide few sources of intermediate demand for the products have a positive though indirect influence of small-scale industries.8 output and employment, on small-scale industrial Moreover, there particularly in the rural areas. Moreover, as is some evidence that, through duction linkages, the large-scale backward pro- firms may have a negative nd in aote S small-scale impact on rural a on luencepoli th industry. Such negative effects small-scale industries will also exan large-scale industry is providing could occur if a protected pilcis the flow scale large quantities of inputs to of uneducated migrants to industries at an unduly high small- These results urban areas. price. In Sierra Leone, the point to the impqrtance scale flourmill, w-,ch provides large- indirect effects of incorporating these the major material input into the evaluation of small-scale bakers, is protected for the grammes. alternative policies or pro by an exclusive import license Such indirect policy effects, thus in and is evaluated however, can only be properly an insulated, monopolistic position. and assessed in an analysis As a result, it has been that explicitly incorporates able to sell its flour to small-scale these intersectoral linkages. bakers at prices over twice the potential import price. 9 that of costs increased markedly Consequently, bakers have and have been placed in a disadvantagedhad their market position. Thus, even 8.2 A REVIEW OF EXISTING when linkages are considered, PROGRAMME AND policies the INSTITUTIONAL designed to influence lEirge-scale SUPPORT industries would generally appear to exert indirectly a negative influence firms. on the small-scale According to the national development plan,2' 1974/75-7879, Government of Sierra Leone's the programme strategy shows that phasis will be given to development em 8.1.6 Agricultural policies of crafts industries and trades with a high growth (I) potential, such as the following: indicateThe empirical and analytical handicrafts such as that agricultural policiesresults generated by gara printing, country small-scale have an important indirectthis study effect also cloth and woodwork industries, on expatriates)which cater to demands of tourists Such effects particularly and which also have exportand local elites (including will occur because thethose agricultural located potential; dustrial sectors andin the small-scale rural in- have been shown to be importantly areas. (2) industries such as another through linked to one manufacture of components and both the income and production (3) trades such as servicing spares, which relationships. In and repair of equipment (automotive, particular, electrical, electronic, etc.), the Sierra Leone study has the use of which has been increasing agricultural incom demonstrated not only that for the es provide on the the prim ary sources of demand rapidly and is products of small-scale industries econoy and isexpected inete tot increasen r d with modernisation incomes but also that agricultural of of the the and the demand for the industrial economy and rise products are strongly and in the standard of living. positively related The principal components agricultural to one another. Indeed, of the programme are: incomes is shown to generate a 10 per cent increase in a 9 per cent increase in demand for small-scale industry the (1) training dition, products in Sierra Leone. and skill formation; there are forward production In ad- (2) extension; since blacksmiths linkages between the sectors, provide the vast majority (3) credit and inputs by of the tools required as marketing; the agricultural (4) formation and industrial linkages that relate agriculturalsector. There are co-operatives. processing alsoto the backward agricultural itself. Consequently, in productionsector The view of the importance of these development plan also recommended and income linkages, production Small that a new Directorate general policies, for example, Industries and Handicrafts of increase agricultural output designed to be created within the Ministry and incomes would, in general, Trade and Industry to organise of appear to specialised training of craftsmen and artisans training centres and at each in provincial headquarters so as 146 Policy Assessment to be close to rural artisans. In addition, the new A Review launch a directorate is to 147 well-organised programme example, the bill excludes firms seminate of industrial exiension to dis- employing fewer than 15 information on new production the definition of small people in techniques, s;ources of raw enterprise and does not introduce materials, credit, marketing, tives for small enterprises any incen management, etc. As suach, as is stipulated for large-scale recommended the recruitment the plan the Development industries in of five Industrial Development, of Industries Act, 19 81-,Despite ficers, three for the provinces Of- the bill represents these weaknesses, to be based in the training a positive step in government two for Freetown. The plan centres, and concrete policy to introduce programme also included the measures for promoting small-scale ment of an establish- It is essential industries. autonomous handicrafts corporation that such a future small-scale emporium and handicrafts industry legislation in Freetown to facilitate recognises the importance of marketing arrangements. attending to the needs of small-scale industries, artisans, cottage industries and handicrafts as was empha 8.2.1 Current sentedsised in in theChapters 1974 Development Plan. The institutional and programme 2, 3 and 4 show empir;cal evidence pre arrangements the importance of in the Sierra Leone economy. these activities It is suggested that actual As of now, none of the recommendations or potential of the 1974 viability rather ment plan docu- promotional than smallness be the criterion for selecting has been implemented, probably purposes. firms for However, due to financial difficulties. new interest seems to have ernment been kindled within the Gov and among donors to establish framework a workable institutional for promoting small-scale industries develop in Sierra Leone and a feasible programme for implementation. framework of an existing Thus, within the UNDP/UNIDO Project, a major ation of the Ministry reorganis of Trade and Industry has taken quently, the small-scale place. Conse industry department of and Industry the Ministry of Trade has been converted into a Division the Division at equal level with of Industrial Planning and Industrial Statistics and the Division of Operations within that Ministry. ment How this new develop will affect the pace and direction industry of promotional efforts in development is yet to small be seen. Furthermore, the Govern ment of Sierra Leone has recently embarked on an industrial sion programme 2 which provides exten for the development of over rural industi.al extension centres 20 by 1986. Already, 50 officers have been trained extension to introduce new projects, production methods, improved better management and marketing in 6 different provincial techniques locations. Notwithstanding the ures, there is urgent above meas need for appropriate legislation clearly define which would small-scale industries and adequate handicrafts, launch an institutional framework for promotional purposes and out line measures that would be implemented. A recent effort Government of Sierra Leone of the to introduce a definite strategy promoting small-scale enterprises of is a bill' entitled 'The Small-Scale Industries Act 1981'. Unfortunately, due to some weaknesses, the bill was not passed by Parliament and is yet under revision. For Summary and Policy Implications 149 The initial capital requirements at the time these small industries were established proved to be quite modest, generally less than Le.100. Approximately 80 per cent of the funds used to establish these indus tries came from personal and family savings, while 90 per cent of the funds required for expansion were reinvested 9Summmary and M profits. Self-financing has plicdy ajor Policy The total labour input varies quite widely from industry and by location, being significantly to industry Imp cations of lower in the villages than the Stud larger urban in the locations. Apprentices and proprietors majority supply the vast 9.1 SUMMARY OF MAJOR The apprenticeshipof the small-scale industry labour. FINDINGS proprietors system is the primary technical from vehicle for providing The evidence trainingoneof andsmall-scale one-half industry. the generated by this study has to five Apprentices serve the industries are a significant revealed that small-scale years, depending component of the industrial industry. The majority of the on the Sierra Leone. sector of proprietor apprentices are also required Small-scale industries account a learning fee for the training to pay the employment for 95 per cent of the received. This fee is levied and approximately 43 per cent more often in rural entire industrial of the value added of the than in urban areas. sector. The vast majority 90 per cent of the proprietors found to of these industries are had themselves previously be located in rural areas. apprentices. None of the served as Indeed, 95 per cent of the proprietors examined in the establishments, 86 per cent any formal vocational study received of the employment, and 75 training and only one-quarter the value added is generated per cent of formal education. possessed any by rural small-scale industries. Moreover, only 17 per cent of locations, however, tailoring In all even the proprietors kept is the single most important a rudimentary set of books blacksrnithing, carpentry, activity; In addition or accounts. baking and gara dyeing to these descriptive insights r importance, but follow tailoring in into the nature of by quite a wide margin. scale industry, the Sierra Leone small There study also sheds light on the are large differences in the determinants of the demand major annual mean output and value- for and supply of small-scale added generated by the various products. With industry types of small-scale industry respect to demand, three principal Leone. In addition, in Sierra for the sources of demand the level of activity fluctuates products of small-scale industry year, although the widely over the have been identified. seasonal patterns differ from major source is that demand The The industry to industry, generated from the incomes magnitude of capital input urban consumers. The of rural and used by the small-scale firms analysis of demand elasticities varies by industry with the mean also the study undertaken in capital stock ranging from reveals that, counter to a key gara dyeing to US$2079 US$89 in Resnick premise of the Hymer and in blacksmithing. Comparative model, the rural income indicate than statistics elasticity of demand for most these amounts of capital are small-industry products is strongly small by international positive; this result thus standards. The composition that the demand for indicates antly of this capital input also these products should be expected by industry. In all the major varies import- strongly to increase capital industries, however, working as rural incomes increase. constitutes a relatively small A second source percentage of the total capital of demand arises from the backward input. production linkages and forward with the agricultural sector. There is also evidence, ages from The backward link however, that this capital agricultural processing activities, utilised. The amount stock is not fully particularly rice milling, of excess capacity differs to agricultural production are industry, ranging from industry to found to be important. In from 25 per cent in gara strong forward linkages addition, dyeing to 41 per cent in are shown to exist from the blacksmithing. industry, which primarily blacksmithing Moreover, the highest produces and services small amount of excess capacity is farm tools, to the found to agricultural sector; the agricultural exist in the rural localities. sector thus provides an important source of intermediate demand for small-scale industrial products. 148 150 Policy Assessment Summary and Policy Implications 151 The third source of demand for small-scale irdustry is that pro- 9.2 MAJOR POLICY vided by the foreign or export sector. IMPLICATIONS international demand exists for The study reveals that a strong the products of Sierra Leone's gara dyeing industry, In view of the strong economic justification for small-scale industries, The analysis of the three major the major policies that either sources of demand for the small- directly,'or indirectly influence small scale industries thus indicates that scale industry deserve attention. The the existing and potential market provision of capital or credit is for the products of small-scale one of the most commonly utilised industry is strong and provides little methods of directly affecting support for the contention small-scale industries. The of Hymer and Resnick that the demand empiricat and analytical findings for and supply of study indicate, however, of the these products will necessarily decline that capital may not necessarily as rural over-riding constraint be the facing small-scale industry in Sierra income and output increase. Leone. The analysis Improved institutional mechanisms of alternative production processes and to increase the access to and the studies have shown the case availability of credit small-scale that the more -traditional' processes within at somewhat higher, though not penalising, industry are not only more labour-intensive each interest rates charged higher labour-capital ratios), (i.e. possess to by the commercial banks, however, would but also possess ".gher output-capital be of value. Simultaneously, direct policies appear ratios than do the more technical skills within designed to improve 'modern' small-scale industrial processes. small-scale industry would be beneficial. The more In view of the 'traditional' processes thus generate overwhelming importance and ubiquity and a larger both a higher output apprenticeship of the amount of employment per unit system in providing technical of the relatively scarce ation training, some consider factor, capital. Consequently, there might be given to upgrading and expanding appears to be no conflict, at least the existing appren in a static sense, between output ticeship system. There is and employment when choosing also evidence to indicate that improved between industrial processes managerial skills with respect in Sierra Leone. to record keeping, savings, reinvest The profit analysis reinforces ment of accumulated profits the findings derived from the and business planning, would enhance proportion analysis. factor the economic viability The small-scale processes and of these enterprises, thus ensuring sessing the highest industries pos- self-sustained future labour-intensities and average capital growth within the small-scale industry also tend productivity Because sector. to generate the highest rates of economic of the high rates of business failures profit. years within tw&or three On growth aspects of Sierra of a firm's establishment, project Leone small-scale industry, our com- designers must be more parative static analysis of 1974 and cautious in selecting firms for 1980 situations shows that both in long-term financial assistance. They terms of number of establishments might want to select, for example, and employments, there have only firms that have been in been positive trends. In existence for two years localities of over 20 000 inhabitants, or more. However, since the pitfalls annual rate of manufacturing an learning are at a maximum of employment growth rate of during the first few years of a cent has been about 6 per project designers firm's life, recorded. In the rural towns might consider alleviating the and villages (localities den firm's financial bur with less than 20 000 inhabitants), at that time, for example by allowing there is clear evidence of subsec- grace periods of two to toral growth patterns particularly three years for the in the food processing and repair repayment of loans. industries. In addition, policy measures may be needed to facilitate the The results of our study process of change in these enterprises. on the innovative behaviour of Sierra Although the Government has Leone entrepreneurs, though only established near Freetown indicative, shows a tendency on an industrial estate in which subsidised the part of entrepreneurs to introduce facilities are made available to changes in products to en- large-scale establishments, no such hance their marketability, to restyle arrangements are being contemplated their various clothing lines and to promote small-scale indus to adopt new packaging methods try. Therefore, policies or projects for products. There was, to a less which create artisan clusters, extent, some tendency to buy new centres or zones, even without equipment. Our bakery case study subsidised facilities, could facilitate showed a tendency towards the the transmission of technological use of improved traditional brick information among small-scale ovens. enterprises. Moreover, the cost of providing information and techni ovns cal assistance would be reduced. Policy measures may also be needed 152 Policy Assessment to introduce other ways of disseminating information product and technological on possible innovations. Since very preneurs had read few of the entre about the machines they may be bought, for example, necessary to raise literacy it access to information standards and thereby sources. improve needed in some cases to foster Finally, adjustment the use of more appropriateassistance labour- intensive techniques. may be Notes and Such adjustment assistance References cial or technical could be of a finan Very importantly,nature. each ent of the three case studies dimension of revealed a differ- tailored to the policy need, PREFACE unique constraintsthus of individualindicating pie, our results subsectors.that policy Forhas exam- seem to highlight the need to be 1. See, for example, to focus on upgrading of David Morawetz, 'Employment dustrialisation in Implications of In Developing Countries: A industry, ensuring of stability Survey', Economic Journal, and improvement in the sewing machine of access to small rental market September 1974, pp. 493-541. machines funds for the purchase Academy of Sciences, Also see Richard S. Eckaus, Appropriate in the tailoring industry, of sewing 2. 1977) p. 115. while in the bread Fo: more details, see Chapter might be a need for upgrading industry there 1, section 1.3, on survey the traditional oven 3. See, for example, E. Wayne methodology. In addition technology, Nafziger: African Capitalism: to these policies dto in Nigerian Entrepreneurship A Case Study scale in (California, scale industries,tont therethere aries designed small- Hoover Institution Press, are several policiestodirectly that inlec 1977); Sayre P. Schatz, Nigerian direct effects on these have importantml-California in- Capitalism (Berkeley, industries. Fiscal policies, and ProductionPress, in1977); and William University of particularly as Developing Countries:F. Steel, Evidence Small-scale reflected in tax York, Praeger from GhanaEmployment (New designed incentive ordinances and Publishers, 1977). specifically for Sierra the import duty structure discriminate Leone large-scale industry, against small-scale tend to 1 INTRODUCTION need to spell out through firms. appropriate legislationThere is therefore tive measures for promoting various fiscalan urgentincen- small-scale industry. See Morawetz, Employment indicates Finally, the study Implication,-of that policies designed Frank, 'Urban Industrialisationsee also incomes generally to increase agricultural unemployment and econoric C.R. have an important positive, output and growth in A sfria' Oxford though indirect, influ- Economic Papers,July 2. The emphasis 1968, pp. 249-273. ence on employment growth on small-scale industries. reflects the heightened concern tion.for equity, See the high unemployment Morawetz, Employment rates and the high Implications. rates of migra 3. 'Large-scale' in these figures reflects the Ministry and includes those firms of Labour's definition details, employing six or more workers. see E. Chuta ard C. Liedholm, For more The role of small-scale industry in employment generationand riral from Sierra development: Initialresearchresults Leone, African Rural Employment Lansing, Paper No. 11 (East 4. Michigan State University, 1975). Derek Byerlee, J. Tommy and Leone: H. Fatoo, Rural-urbanmigration Determinants and policy implications, in Sierra Paper No. 13 (East African Rural Econcomy Lansing, Michigan State University, 5. Government of Sierra Leone, 1976). National development plan, 197&-79 (Freetown. Government 1974-75 to 6. Ibid., p. 185. Printer, 1974). 7. The results of another Sierra Leone study similar this volume are presented to the ones reported in by D.A. Fowler, 'The informal Freetown: Opportunities for sector in self-empioyment', in S.V. Sethuraman 153 154 Notes and References Notes (ed.), The and References urban informalsector in developing 155 p overty and environmen t (G countries:Employment, e n ev a, IL O , 19 8 1). gricultural processing A g cu t r l p o e is also cx-:luded since s n is a s ex ud d i ce it t isis analysedna y in a separate 8. povent tyGenndva,nviILonOm19 1).study. d i n a e a a e environment,See Chapter 8 for a more detailed For a more extensive 9. See Morawetz, assessment of the economic policy discussion -of*the industry definition, 'see Employment Implications, for Chuta and Liedhoun, the employment-output an excellent discussion of 2. Government of Sierra RoleLeone, of Small-scale lndustry, trade-off. The World Employment 3. Forme om iscusion National Pevelopmentp. 8.Plan, of the ILO has recently stressed Conference 3. For oftheonall-velop nt p. 27. the need for choosing employment a more complete discussion of en ition given used in this the mall-scale industry definition and such a trade-off. For more details, study, see Chuta ard Liedholm, basic needs: A one-worldproblem see ILO, Employment, growth Role of Small-scale Indus 10. The only 'Geneva, 1976). 4Dt p. 9. comparable survey was the study 4. Data obtained from Ghana: Area of small-scale industries in the employment list of the Ministry Ghat,.i undertakensampling it. of small manufacturing December 1973 and supplemented of Labour for 1963. For more details, establishments,see Government 193 by data on eight large-scale of 5. Theseincludsed onsthits firms not (Acra, Central Bureau of estimates liswere t.obtained ie byymliligt -multiplying Statistics, 1965). In Nigeria, th actualata numberubro of (AcrCeResearch Unin of the University the Industrial industrial tra BurautheaUnivestics,of of Ife19 has undertaken establishment establishments and industrial I nNiertenstial n tion of that employment in each of the 24 enum eration in several states, sample village'enumeration 'enumeration area' in that particular but does not estim ate establishments areas' by a figure reflectingagricu*Zural region. It smaller villages (below 2000). For more details, see Industrial Researchinsh ud the representa Unit, Small-scaleindustries: Western should beb furtherf rhe notedno d thatt ati individuals iv u engaged State of Nigeria(Ile-Ife, University ls ng ed inni industrial d sral activitiesciiis of Ife, 1972). on a part-time Liedholm, For a description of other partial basis are included in these Research on Employment in surveys, see Carl data. Africa, African the Ru~l Non-farm Sector in 6. Computed from Table 2.1. Rural Employment Paper #5 7. The average size of (East Lansing, Michigan firm, in terms of the number employed, State University, 1973) p. 6. Freetown (3.5 workers) than is larger in in the enumeration areas (1.6 8. The breakdown workers). 11. See ibid., p. 4.below.11 eeiidp 4. of the individual mean value 12. The 1963 population census of Sierra added is described in refer to cities, towns and Leone used the term 'locality' to detaleo = US$1. 13. The only villages. 9. 1Le(leone) oSierrAl0during the 1974-75 survey period. For more constraint imposed on the random see Government ingthe174-7 surferd Leone,details, that at least one selection procedure was of Sierra Leone: Natonalaccounts locality had to be selected from to 1971/72 (Freetown, Central of SierraLeone, 1963/64 agricultural regions that each of the eight Office of Statistics, 1973), p. 31. were used as the sampling frame 10. S.A. Aluko, Employment in . the 'enumeration areas'. This for choosing sented Nigeriansmall-scale industrialsector, localities representing all typesprocedure was designed to ensure at Ford Foundation Conference pre of agricultural production conditionsAthat ment, on Technology and er Delhi, Idi, a1973, and Employ were included. ni da o u 'C r ct rsJanet cs Faroog, Ghazi M. Faroog nd mp y e t g n r t onand 14. Each 'enumeration small-scale f area', which was an artificial grouping industries in Western Nigeria', the Central Statistical Office constricted by in The PakistanDevelopment estimated for the 1963 population census, Review, Vol. XV No. 4 1976, 200 families. For more details, see contains an Staff pp. 382-405. farm management Dunstan Spencer, Micro-level Working Paper No. 191 (Washington, and production economies 12. International Labour DC, 1974). African farmers: research among traditional Office: 'Sharing in development: AeranNo.ames Lessons from Sierra Leone A programme of Paper, No. 3 Lsnsng", ierraig oneAfricanAfrican Rural Employment (East Lansing, Michigan State University, ural Em t employment, equity g 15. The 1972). 13. R.S. Berry, The relevanceand growth in the Philippines' 96 establishments selected from Colombia, Discussion and prospects of small-scale(Geneva, 1974). the 24 village 'enumeration areas' Paper No. 142 (New Haven, Connecticut,industry in were all selectr-d omic Growth Econ 16. facturing,For more 1973-.74details, randomly. see(Freetown, (See 14). Centre, Yale University, 1972). GovernmentCentral of SierraPlanning Leone, Unit). Census of Manu- 14. andFor oraJohn review M. of' these studies, survey, World Page,age , Jr., SmallSeerisee enterprises Liedholm, in Research on Bank Staff Working Paper seArcnAfrican development:eeloment;employment; A No. 363 (World Bank, 2. DESCRIPTIVE PROFILE Washing OF SMALL-SCALE INDUSTRY IN 15. SIERRA Chuta and Liedholm, Role of small-scale LEONE industry, and Tables 2.3 and 1. The "industries' 2.4. examined in this study include 16. Computed from Table 2.1. ments that specifically only those establish- engage in the production 17.1 8 Computeda taproi ma thfrom 3 Table tured goods' and this excluded and repair of 'manufac- er c nt 2.4.e inRuraldit e t small-scale industries those establishments engaged 18. Data from other countries at lruaindustr a li ad account for in mining, significant indicate that rural industrial employmente mp o me t i in many other countries and is construction, trading, transport, often exceeds the total number financial, social and personal services. employed in all urban industrial firms. See E. Chuta and Carl Liedholm, 156 Notes and References Rural oeaeloyment aeviewo Notes and References h S tateoeM.rU Rural D 34- The 20 per cent Eve lopmentP Apeo.:4(EastLaigh discount rate was also used 19. Some of the Statega to convert directly the data presented in this section e AniSv~eur..U working apprentices about the characteristics capital stock into a figure and entrepreneurs, of 35. Unfortunately, reflecting the opportunity however, refer to it has not been cost of funds tied up both randomly and possible to undertake analyses in Purposively sampled Frrms. tai the 'optimum' to asce 20. dueIna fewlocaities These are noted in the inventory levels for these to difficulties with'frexartple, the enumerators. several respective tables. firms. difficulties in monthsIn the enumeration ofdata could not be 36. See Part II for a description identifying some of the areas, therecollected were carpenters, of these more 21. The gross output non-farm householdsj thus, those gara dyers or bakers 'modern' firms. No 'modernm value of each sample firm, capita! figures reflect morewere included in xhe random ' by specifying that all for example, was obtained 37. For more details, 'traditiond processes. sample; the output, including that see F.C. Child, Ane mpirical barter or gifts, should produced for inventory, industry in Kenya, IDS study of small-scale be valued at the Development Working Paper No. rural material sales price at the firm. Studies, University 127 (Nairobi, Institute and services input value The of Nairobi, of value of all for each firm reflects Steel, Small-scale employment 1973), and William F. indirect raw materials, lubricants, the purchased andproduction taxes and telephone fuels (including Evidence from in developing countries: services consumed by electric-.y), water, Ghana (New York, Praeger, during the year. the sampled firm 38. 1977). The value added for each 39. SimilarChild, Empiricalproportions firm is simply the resulting Study.are reported elsewhere difference between Enterprises in Africa by Page, Small the value of gross output in African Development. 22. *Gara'refers to and the value of material the plani from which the 40. For o dye is obtained a good discussion of the For more Economic capacity utilisation issues, details, see Chapter 6. Literature,Dec. 1974, see Gordon 24.23. ComputedCmuefrmfr-eesuvydt.*factors. pp. 1301-1320. Excess 25. from farm-level survey capacity can be datTheseCollted figures andrifa-level subsequent data. both planned and unplanned surveyimport dt. data werer41. and caused by data collected during phase obtained from the input both demand and supply 26. Government 2 of the small-scale industrial capacity,For more details on of Sierra Leone, National survey. see N Phan-Thuythe use of (ed.) time-intensity approach 27. Although Development Plan, p. promoion (Westead Industrialcapacityto and measure excess electricity is available 190. 42. England, Gower employment in virtually all localities These data were obtained Publishi it, 1981), pp. with with 5000 or from the labour input 22-4. 20 000 inhabitants or less sample,administered twice questionnaire that was used electricity. a few gara weekly to firms over the 28. This definition dyers actually period of one year. clude cash net is somewhat more 43. It should be noted, used apprentices receivables. For small-scalerestrictive than definitions however, that in the in production, establishments, however,that in- larger random and: purposivebut 30. cFiraoesdre net 44. their utilisation was receivables are minimal See D.S.C. Spencer very minimal. inimal~and cash data are and Derek Byerlee, Technical 29. Computed difficult to obtain, change, labouruse 30. Fora from survey Ynoahdtaedfutto and capitalgood review ofthis issue,data. see small-farmer development: inputs for agricultural productionP.A. Yotopoulous, 'Fromoban stock Evidence from Sierra functions: A micro-economictoflow Rural Leone African 31. approach', in Journal 45. R rlEEconomy Working Paper Not unexpectedly, theof mean Farm Economics, It should obeonotedy W ri g No. 15 (East Lansing, Michigan widely, expected lifeMay for each1967, pp. tha, inP purbanrN .1aias,E the State ranging from 2 years for component varied 46.pComputedifomJdataacolletedrfromotheietrepreneurialJep.ionnaires aproprietors sn ,M ci generally for the blacksmiths' non-metal tools used by 46.45.91ppestionnaires.ComputedItr o dbe ofrom and data o a tt anvils. These results tailors to 35 years dcollected rthi fromrba thea~ entrepreneurial adjusting the capital thus reflected the 47. For an excellent teproprietos g neal data for life expectancy. importance of discussion of the influence A similar procedure was factors on entrepreneurial of social and Psychological neurship perfor luancesee John used to calculate the cnd economic development, R. Harris, Entrepre 32. establishments. mean expected life of Working Paper For a detailed discussion those buildings owned by No. 69 (Cam McKinnon, of 'fragmented' capital the bridge, Mone and market, see R.I. For example,Massachusetts Massachusetts DC, Brookings capitalin economic tionships withinattitudes towards risk Institute of Technology, 33. institution, 1973). dcvelopment (Washington, an organisation canard modes of inter-personal1971). Linsenmeyer, for example, be shaped bv these social rela reports that fishermen 48. psychological and fixed capital at a rate adjusted obtained loans Callaway foundvariables. that for default risk for were 48 per cent of the proprietors' details, see Dean Linsenmeyer: of 43 per cent For more farmers, while Harris fathers in Ibadan nologies Economic anclysis fathers reported that 25 per cent in Sierra Leone of alternativetech- were farmers. For more of his proprietors' Program marine fisheries, African tional crafts details, see A. Callaway: Working Paper No. 18 Rural Economy to modern industries', 'From tradi (East Lansing, Michigan on itsstructure in The city of Ibadan: State Univer- anddevelopnnt, edited A symposium and B. Awe (Cambridge by P.C. Lloyd, A.M. sity, 1976). University Press, Mabogiuje 1967) and Harris, Entrepre neurship and economic development. 158 Notes and References Notes and References 159 49. For results from other African countries, see Callaway, 'Traditional 11. Computed from sales data collected as part of the small-scale industrial Crafts to Modern Industries'. 12. Computedsurvey. DETERMINANTS from input data collected, by 3.ADETE OF DEMAND FOR AND SUPPLy farm-level study, for more SMALL-SCALELEINDUSTRY OUTPUT AND EMPLOYMENTOF details, see Chapter 13. Shcvels were the only1, section 1.3.1. Cutlasses and knives weretool not prodced in large quantities locally. also imported but they represented less than 10 per cent of the total 1. See Stephen consumption. In 1973, Hymer and Stephen Resnick: 'A model matchets for example, only of an agrarian 1.4 million.were imported, while domestic 129 000 economy with non-agricultural activities', production was approximately in The American Economic 14. Review, Vol. 59 pp. 493-506. See also Computed from initial output data generated and Howard Pack: 'Unemployment by farm-level study. The income distribution in Kenya' in Economic 0.01 input-output coefficient could be used Development and future demands, for making projections of CulturalChange, Vol. 26 however, only if the marginal and average 2. For more details, pp. 157-8. coefficients see R.P. King and Derek Byerlee: 'Factor intensities were similar. and locational linkages of 15. Donald L. Huddle rural consumption patterns in Sierra Leone', and Thi-Min Ho, The contributionof traditional in American Journalof small-scale culturalgoods and AgriculturalEconomics, Vol. 60 pp. 197-206. in internationaltradeand employment, Paper 3. The subsistence ratio No. 35, Program of Development Studies (Houston, difference was defined as subsistence consumption (i.e. Texas, Rice Uni between what a household produced and sold) versity, 1972). of total consumption as a per cent 16. Hymer and Resnick, and was included as an independent variable Model of an Agrarian Economy. because a household's pattern of consumption 17. The data presented in the previous chapter tion towards is related to its orienta- activities on the growth of these the consumption of home-produced goods. since 1974 reinforces these demand conclusions. 4. A coefficient 18. These variations of 0.85 indicate" that, if total consumption by location were statistically significant at the increases by 10 per expenditure cent level 5 per cent. theni the demand for small-scale industry on the basis of Chi-square tests. products would increase by 19. Robert Dorfman: 'Mathematical 8.5 per cent. The expenditure elasticity for or "Linear" programming: A non large-scale industry mathematical exposition', in American higher, products at the mean income level is Economic Review, Dec. 1953, 1.02, than that for small-scale industry products. only slightly pp. 797-825. 5. Similar results 20. The full array were obtained when the log-log inverse of processes for the tailoring kclothing), gara.and (i.e. In C industries, however, breaa In A + b, In Y + bjY), as weil as will be examined in the empirical casejfltudies to estimate the total expenditure other functional forms, were used of elasticities. The terms are the same as Part II. those previously defined for the semi-log 21. The more sophisticated carpentry equipment, elasticities model. Although the income such as the Italian electri can vary by income class in many of cal combination machines, cost about were no statistically these formulations, there 22. This Le.4000. significant variations in the elasticities modem equipment costs from Le.30 000 to levels. by income 23. See Morawetz, Le.60 000. Thus, for the range of incomes included Employment Implications for a good review income elasticities in the survey, the relevant of the for small-scale industry did not decline significantly issues surrounding this use of the ratio. as income levels increased. 24. See, for example, Steel See King and Byerlee, 'Factor Intensities who uses the ratios for this purpose in Ghana. William F. Steel, Urban unemployment and and Local Linkages'. small-scale enterprises in 6. Our results are based on a cross-section must be cautious study of households and thus one developing countries:Evidence from Ghana about making judgements about the behaviour 25. See Child, EmpiricalStudy. (New York, Praeger, 1976). sumers over time, particularly of con- 26. To facilitate international 7. in the long run. comparisons, however, these ratios have also Spencer and Byerlee, Technical Change, been computed in terms of Labor Use, and Small Farm 27. input stocks and are presented below. Development. The assumption that labour is a surplus resource 8. Sierra is somewhat true in Indeed, except for illegal imports of flour from Leone where there is considerable seasonal flour used Guinea or Liberia, all the rural unemployment in by the small-scale firms comes from the large-scale areas and a substantial pool of urban labour. Freetown. mill in resources besides If labour or other capital are scarce, then they should be included measure of total scarce in a 9. Some carpentry products are used by large-scale industrial firms. factor productivity rather than simply capital 10. productivity (see Moraetz, Employment Implications). efficiencyDunstan Spener, Ibi 1. May-Parker and Frank and incomes in the rice-processingindustryRose, Employment, 28. of Sierra Leone, 29. TheseSee ibid for a recent discussion of the employment-output African Rural Economy Working Paper results are subject to the same qualifications trade-off. gan No. 13 (East Lansing, Michi- that apply to State University, 1976). analyses assuming that capital is the only 30. The relationship scarce resource. between processes, factor intensities and economic 160 Notes and References Notes and References 161 profitability will be examined below. These results based on these factor per hour, carpenters at Le.0.35 per hour (assunled intensity measures will be misleading should other inputs other than to be same as tailors in the absence of a good estimate), blacksmiths at Le.0.30 per hour and capital prove to be binding constraints. bakers at Le.0.54 per hour. For more details, see Liedholm and Chuta, 31. The 'modem' processes were not examined by location because there 1976, Economics of ruraland was an insufficient number of observations by urban small-scale industriespp. 68-76. location. 45. At7 the 10 32. The number of workers' data were obtained per cent significance level. l from survey data, while the 45. Ate 1per cent significancenlevl capital stock data were obtained from Table 2.7. 46. See Spencer and Byerlee, Technical C6ange, Labor Use and Small 33. The capital stock, it should be recalled, is measured in terms of its original cost. 34. wasper Farm Development. 47. Analysescent level.of variance indicated these were different statistically at the 1 34. The large-scale industry coefficient was computed from the primary 48. 5 per cent level. data supplied by the large firms to the Economic Planning Unit. The 49. Indeed, capital figure used is the original undepreciated value of fixed assets and the simple correlation coefficient relating the economic profit rate to the labour-capital ratio was +0.50 and was significant at the does not include working capital: thus the capital-labour ratio 0.1 large-scale industry is somewhat understated. for per cent level. The ratio for individual 50. See Kilby, African Enterprise firms ranged from US$66 and Harris, Entrepreneurshipand Econ 000 to US$00. omic 35. See Morawetz, Employment Development. Implications p. 526, for a review of the 51. See ibid. evidence. 5 . Se ibe . id 36. See ILO, and Steel, Urban Employment and small-scaleenterprises. 52. For more details, see Kilby, 37. Steel, ibid, and Child, Empirical African Enterprise, and Harris, ibid. Study. 53. See ibid. 38. The large-scale industry ratio was calculated from the primary data 55. Indetd, the causation may be reversed, since more profitable firms have supplied by firms for 1974 to capital ratio is overstated to thethe extentEconomic Planning Unit. The output- more funds to reinvest. that working capital is excluded not necessarily a causation.A statistical association has been established but from the capital stock measure. 39. See Morawetz, Employment Rural Implications and Chuta and Liedholm, Non-Farm Employment, International Labour Office, Sharing in 4. GROWTH ASPECTS OF SMALL-SCALE INDUSTRY Development: A Programme of Employment, Equity, and Growth the Philippines(Geneva, International Labour Office, 1974), for for a list of 1. For example, see A. Nag, 'Small industries: these studies. The evidence from the Philippines, ILO, Aspects of their fiiortallty', Sharing in in Economic Times (Delhi) 6 Oct. 1979 Development p. 541, indicates and Albert Berry and Armando with firm size. that the output-capital ratio increases EmploymentPinell-Silas, Small-scale and Rural enterprisesDevelopmentin Colombia:A case study Studies in 40. No. 56 (Washington, DC, These aggregate comparisons must be used with great caution for policy World purposes since the product-mix differs significantly between Bank, 1979). large- and 2. In half the villages, for example, emviployment small-scale industry. The most appropriate comparison is in manufacturing in between firms creased, while in the other half it decreased. producing the same product with the same degree of vertical integra- Depending on the aggrega tion procedures used, the rate of change ranged between plus tion. See Chuta and Liedholm, 1979, Rural Non-Farm Employment. oi minus The three case studies one per cent. in Part II below provide such comparisons. 3. How 41. See above for a discussion of the opportunity do these aggregate employment and enterprise growth rates in cost of capital. Sierra Leone compare 42. Thus, based on the previous marginal productivity with those in other countries? Unfortunately, analysis, tailoring such dynamic data are generally apprentices were valued at Le.0.08 per hour worked, blacksmithing available for only a few countries. In apprentices at Le.0.05 the Philippines, Anderson and Khambata report that, between 1967 per hour worked, and gara, carpentry and baking and apprentices at Le.0.08 per 1975, employment in household and small-scale manufacturing hour (assumed to be the same as tailors in the establishments absence of a good estimate). See with fewer than ten workers increased at an annual rate Liedholm and Chuta, 1976, Economics of 1.9 per cent. of Rural and Urban Small-Scale Industry, African The growth rate of employment in these was found to be Rural Economy Paper higher than that of all categories No. 14 (East Lansing, Michigan State University, 1976), p. 70. of establishments employing ten 43. The economic profit must persons or more. Moreover, as in Sierra Leone, the increase be kept distinct from the accountant's con- manufacturing in urban cept of profit which usually employment exceeded that in rural manufacturing. Be includes the opportunity cost of the return tween on capital and proprietor's labour. 1960 and 1975, for example, household and small-scale manufac turing (fewer than ten workers) 44. Thus, based on previous marginal productivity analysis, tailoring in Manila and Rizal increased at a 6.6 pro- per cent annual rate, while elsewhere the prietors were valued at Le.0.35 per hour, gara proprietors at Le.0.84 annual rate was only 1.6 per cent. See Dennis Anderson and Farida Khambata, Small enterprises 162 Notes and References Notes and References 163 and development policy in the Philippines: A case study World Bank 2. Peter Kilby points out that the cotton Staff Working Paper No. 468 (Washington, DC, textile ifidustry has been one of 1981). In Lomd, Togo, the major areas of industrialisation for the number of manufacturing establishments some developing countries in the was estimated to grow at 5 iwentieth century. See Peter per cent a year. See Georges Kilby, Industrialisationin an open econ Nihan, Erik Demol aid Comlavi Jondoh: omy: Nigeria, 1945-66 'The (Cambridge University Press, 1969) p. 108. modem informa sector in Lomd', in InternationalLabourReview 3. Harry G. Johnson, Economicpolicies tdIyards less developed countries Sept.-Oct. p. 633. (New York, Praeger, 1968) p. 74. 4. See Franklyn Lisk and Rolph van der Hoeven. 'Measurement and 4. The aggregate number of such firms was obtained through a detailed interpretation of poverty in Sierra Leone', in International Labour street-by-street enumeration of small-scale industries in Sierra Leone. For Review Nov.-Dec. 1979, p. 727. details, see E. Chuta 5. See Government and Carl Liedholm, Role of small-scale industry. of Sierra Leone, National accounts of Sierra Leone, 5. The use of the representative 1974-75 to 1977-78 (Freetown Central Statistics Office, firms in the microeconomic analysis of an 1979). industry has the advantage of enabling us 6. See Lisk and Van der Hoeven, op. to examine the decision cit. making behaviour of a manufacturing 7. Liedholm and Chuta, The economics cf enterprise type without necesrarily rural and urban industries, focusing on every firm within the 8. In Freetown, for example, industry. For the purpose of these case the 'average' size of a small enterprise studies, representative increased from 3.2 to 3.8 individuals, firms have been delineated according to the while in the smallest villages it observed technique of production increased from. 1.9 to 2.2. and sometimes with respect to the location of firms as well as the quality of inputs. 9. See Enyinna Chuta, Choice of appropriatetechnique in the African For more details see A.E. Buckwell and P.B.R. Hazell: 'Implications of aggregation bias for the bread industry with special reference to Sierra Leone (ILO, Geneva, construction of static and dynamic linear programming supply models', 1981: mimeographed World Employment Programme research in working Journalof AgriculturalEconomics, Vol. 29 No. 2, 1972. paper; restricted), pp. 12-16. 6. A firm is considered technically 10. On the basis of the total listing of enterprises in each sample inefficient if it utilises more resources or locality, a equal amounts of some resources and more of others sub-sample of every fifth firm (giving a total than any other firm of 307) was systematically to produce a given amount of output selected for more detailed questioning. at the existing level of capacity utilisation. The reciprocals of the factor proportions 11. Ten firms which said that they had reinvested some (shown in parenthesis of their profits could in the relevant Tables) show how much capital not remember the exact amount: and labour are required to 67 of the 128 sampled firms did produce one unit of value-added. reinvest. 12. At the time of the survey in April 1980, 7. For more details on the theory of asset fixity, see Glenn L. Johnison ahd US$1 = Le. 1.05. C.L. Quance (eds), The overproduction 13. These compound rates can easily trap in US agriculture:A study be verified by using the following of resource allocationfrom equation: K79/80 = - World War Ito the late 1960s (Baltimore, The K75/76 (I + r) where K = capital stock; r = rate Johns Hopkins of reinvestment. University Press. 1972). Also see Harvey Leibenstein, General efficiency theory and economic 14. In order to compute the real rate of development (New York, Ox investment, we deflated the nominal ford University Press, 1978). five-year reinvestments by 212. which is the national consumer price index 8. Economic efficiency means the ability of firms to generate for 1978 and the mid-point bet.veen 1975-76 and 1979-80. both higher See United value-added and employment per unit of scarce Nations: Monthly Bulletin capital, particularly in of Statistics (New York), Vol. XXXV, No. 9, the short run. Sep. 1981. The consumer price index is the only price index available in 9. Total clothing demand can be defined as the sum of the Sierra Leone. See also International Monetary Fund: Surveys values of of African clothing imports, gross outputs of the two large economies (Washington, DC 1975), Vol. 6, pp. 4109-413. firms, and the small scale tailoring firms. The relative shares in total demand are 13 per cent 15. See Liedholm and Chuta, 1976, Economics of rural and urban indus- imports, 7 per cent for the two large-scale firms, and 79 per cent for the tries, p. 94. small-scale subsector. See Chuta, Linear ProgrammingAnalysis. 16. See, for example, Gustav Ranis: Industrial efficiercy and economic growth: A case study of Karachi (Karachi, Economics, 1961). Institute of Development 6. THE GARA (TIE-DYE) INDUSTRY 6 H A A ( I - Y ) I D S R 5TEconomics, 19 . I1. See Chapter 2, 5. THE CLOTHING n22, for the meaning of 'gara'. INDUSTRY 2. See Chapter 2. 3. See Liedholm and Chuta, 1976, Economics of rural and small-scale 1. In terms of number of establishments, cloth working appears to be the industries. most important small-scale industrial activity ,iuroughour Africa. For 4. Ibid. more details, see Liedholm, Research on empcyrnent, p. 7. 5. Chuta, 1977, Linear Programminganalysis. 164 Notes and References Notes and References 165 6. G. Perkin and A.E. Everest: The naturalorganiccolouring matters. For C.l.F.) of synthetic organic dyestuffs and caustic a more detailed treatment of this topic, see E. Chuta and soda (sodium hydrox M. Steacy: A ide) have risen at annual rates of 16 per cent and case study of garatie-dye technology in Sierra Leone 4.4 per cent respec (Njala University tively between the period 1963 and 1973 (Government College, 1975; mimeographed). Quarterlytradestatistics, 1963-73, Freetow, Central Statisticsof Sierra Office).Leone: 7. G. Perkin and A.E. Everest: The naturalorganic colouring matters 20. The quoted prices are relevant for the 19-75 survey (London, Longman, Green, 1918). period. 'Santos', 8.LSomet , eLegan e rh 20 h utdpie aerlvn o h 19~-5sre eid Sno' 8. Sometimes, the Lebanese n ts s e t'Omega' and 'Super' brands were sold in widths of 50 inches to 52 erchants store the gara in sacks weighing 40 inches, satin in 36-inch widths, lbs and gara workers purchased poplin in 31-inch to 32-inch widths and the gara in that form at a price range of shirting in 27-inch to 28-inch widths. Le.6 (Le.1 = USot.10) to Le.12 (at the time of the survey), depending 21. Only 10 on location. out of 22 selected firms had reliable input-output data for a 12-month period. 9. Information concerning the origin of gara craft was supplied by Mr. 22. Given an average fabric production of 2083 linear yds with a standard M.A. Tunis, Head of the Small-Scale Industry Division, Ministry of Trade and Industry, Freetown. Christopher Fyfe (1962) contends deviation of 1906 and a population of 360 gara firms, an upper con gara cloth that fidence limit of fabric production is about 1 504 was not produced in the 1850s when the Madirgos and Susu approximately 000 linear yds (or first migrated into 1 880 550 sq yds) where n = 10 and ta/2 or t .005 = 3.3. Sierra Leone. The trade exhibition of 28 February Even the 1865, which took place linear programming prediction of Le. 1079 091 of gara output in Water Street, Freetown, Fyfe, referred to the for 1974-5 in sales of country cloth rather Sierra Leone converts into 599 495 linear yds (or 749 369 than gara. See Christopher Fyfe, A history sq yds) of fabric, assuming ofSierraLeone (Oxford University Press, a weighted average output-fabric ratio of 1962). Thando Gwebu points Le. 1.8 (see Chuta, Linearprogramming out that in the early 1920s, Chief of Rogbane allocated analysis). the central part 23. Liedholm and Chuta, Economics of rural of Makeni called Manikala to the Madingo migrants, who were skilled and urban industries, p. 54. 24. The bulk of the gara cloth produced in Sierra Leone originates gara dyers. This area still remains the centre of gara production in from Makeni (see private small-scale entrepreneurs who are scattered over the country. Thando Gwebu, A study ofgara industryin Makeni (Njala However, University College, 1971). production is also carried out by co-operatives. At least five such co-operatives or voluntary associations 10. Another traditional dye used in Sierra were identified during the Leone is the kola nut (cola 1974-5 survey of small-scale industries. acuminata)dye. This dye is obtained by pounding kola nuts in a mortar, 25. Some of this class of dyers are in possession of sources of mixing the pounded kola nut with water, and sifting the concentrated information, dye solution. such as Ann Maile, Tie-and-dye. When a fabric is dyed with the native indigo and kola nut 26. While dye solutions, the the average price per lappa for firm types A and B ranges resultant colours after dye migration are dark green between Le.2 or dark brown shades. and Le.3 and Le.3 and Le.3.50 respectively, firm type D 11. Lappas mean the same as wrappers. sells each lappa for Le. 10 to Le. 12. One lappa is equivalent to about 27. In fact, most gara dyers two yards of cloth which may be of in the D category receive orders for gara cloth varying widths, by telephone. Thus, the market 12. For more details, for their output is guaranteed, even at see Chuta and Steacy, Garatechnology, higher prices. 13. For more details on the various patterns, see, for example, Ann Maile, 28. Complete input-output information was available Tie-and-dye as a present-day craft (Lon in, Mills and Boon, 1965). for only one out of the two co-operative societies that were managed by expatriate (mis 14. About 80 per cent of gara dyeing proprietors are women while over 70 per cent of total labou. input is female. sionary) personnel. 29. Such close contact could be in the form of official supervision, 15. For more details, see Chuta and Steacy, Garatechnology. prior to 16. The amount of tying time depends official registration. not only on the pattern desired but 31). Don Jordan, after his surveys importantly on the type of fabric. It of European and Australian markets, will take an individual about 15 recommended some export production minutes to tie a lappa of some cotton fabrics, but specifications like longer fabric one individual will lengths, wider variety of basic fabrics, greater require a maximum of one day working ten hours . tie patterns flexibility of usage, on a improved quality and more distinctive patterns of colours dozen lappas of satin fzbric. The cost of tying by hand for instanIce (see Don Jordan, 'Small-scale production' which way?', in Small IndustryBulletin varies from Le.0.20 to Le.2 for one lappa depending on the complexity of the pattern for Asia and FarEast Vol. 9, 1972. to be tied. 31. The 17. Such drums were uncertainty of future expansion of foreign markets for handicrafts purchased second-hand at a cost of about Le.4 has been pointed ($4.40). The number of drums used out by some researchers. For example, see Maude depends on the volume of activity. Wahlman and Enyinna Chuta, 18. See Chapter 2 for our 'va!vuation procedure. 'Sierra Leone resist-dyed textiles', in The fabric of culture edited by Justice Cordwell 19. An ud valorem customs duty of about 36.5 and Ronald Schwarz per cent is imposed on the (Paris, Mouton Press, 1979) pp. 447-466. ir:portation of these ingredients. Also, the prices per cwt (Freetown, 166 Notes and References Notes and References 167 7. TT-UE BREAD-MAKiNG INDUSTRY 20. See Kilby, 1965, African enterprise,p. 38. 21. See Chapter 2. 1. For more details, see Enyinna Chuta, Choice of appropriatetechnique in the African bread industry wi;z special reference (ILO, Geneva, Jan. 1981; mimeographed; to Sierra Leone World Employment Pro- 8. A REVIEW OF POLICY El gramme research working paper; restricted). V UM rKU.ilKA ME AND INSTITUTIONAL SUPPORT 2. E.J. Pyler: Baking Science and Technology (Chicago, Siebel Publishing Company, 1973), p. 1124. 1. See Chapter 3. A peel is a longhandled spade-shaped instrumen! 2. for getting something used chiefly by bakers 2. Based on the survey data. in and out of the oven. 3. 4. Pyler, Baking science, There was also no direct indication that working capital was a binding constraint. Working 5. Most traditional bakeries revealed a product capital represented, for example, a relatively small damage of around 4 per percentage of the capital cent due to poor oven conditions. of even the most economically viable small 6. The scale industries. distribution of firms is based on the extrapolation of figures ob- tained from the analysis of sample 4. The Credit Guarantee Scheme was launched on I ;anuary 1974, data. during the first and 7. Sponge fermentation is the leavening year 155 loans amounting to Le.520 000 (1 Le. = US$1.20) of dough by a starter produced were guaranteed. Only from spontaneous fermentation. For more technical seven of the guaranteed loans, however, were for details, see Manuel industrial activities. These data A. Matz, Bakery technology were supplied by officials of the Bank of and engineering (Chicago, the AVI Pub- Sierra lishing Company, 1960), p. 377, and Edmond B. Bennion, Breadmak. Leone. ing: 5. See National Development Bank, Loan scheme for Its principles andpractice (Oxford University Press, 1967) p. 100. small borrowers, a The bulk of sponge brochure (Freetown, Atlantic Printers Ltd., rno date). fermentation is now being replaced by the Chorley 6. Bread Process of mechanical J.D. Rogers and I.I. May-Parker: The role of credit and financial development of dough. See FAO, Compos- institutions itc FlourProgramme, Documentation in rural development in Sierra Leone', in National work Package Vol. 1, 1973, p. 77. shop on 8. See Matz, Bakery Technology, agriculturalcredit and banking in Sierra Leone (Freetown, p. 3. Bank of Sierra 9. This point has been emphasised by some researchers. For Leone, 8 Oct.-4 Nov. 1978) pp. 12-30. example, see 7. As shown in Chuta, 1977, A LinearProgramming, Peter Morris and Anthony Somerset, The African Entrepreneur changing the oppor (New tunity cost of capital does alter the intensities York, African Publishing Co-poration, 1971). of the 'processeS' tfiat 10. This definition is somewhat would operate. more restrictive than definitions that in- 8. Indeed, clude cash and net receivables. only about 2 per cent of the secondary-school students are 11. See Chapter 2. enrolled in a formal technical or vocational training institute. 12. 9. A thesis written on Nigeria by Mabawonku attempted See Liedholm and Chuta, Economics of rural and urban industries, to quantify the costs pp. 52-4. and benefits of alternative methods of small-scale industry training. 10. There 13. For more details, see Chapter 2. are several other policies that directly or indirectly affect small scale industries, such as the 14. The available case studies of bread making in Africa provision of research on trading services, include Peter Kilby, government procurement or export African enterprise: The Nigerian breadindustry (Stanford, Hoover Institu- promotion or licensing schemes. They have not been explicitly dealt with at this time due tion Studies, 1965); S.M. Wangwe, Technologicalchoice and technological either to a lack change of data or "uncompletedanalyses. in the bakingindustry- the case of Tanzania(Department of Econ- omics, University of Dar 11. This provision appears to be without time limit. es Salaam, 1980; mimeograph; R. Kaplinsky: 12. Since 'Technological choices in bread making small firms generally are able to escape income taxation, the in Kenya', in Technology, employ- income tax exemption ment and basic needs in foodprocessing does not appear to cause undue distortion in the in developing countries edited by relative post-tax Christopher G. Baron (Oxford, Pergamon Press, 1980). profitability of small and large firms. 13. Revenue and development objectives, 15. See Kilby, 1965, African Enterprise. however, need not be in conflict 16. Ibid., p. 61. The exact figures are 2 with one another. to 5 shillings per bag for wood-firing 14. The duty on some znd 6 to 8 shillings per bag for oil-firing ovens. of the final goods potentially competing with the 17. The studies small-scale industry products, however, tends to be rather of Kilby and Wangwe did not disaggregate their results for low. Small urban and rural locations. agricultural tools, for example, are taxed at only 2.75 per cent. 15. Cloth duties are specific rather than ad 18. See Kaplinsky, Technologicalchoices, p. 280. valorem. The 22 per cent figure 19. Ibid., p. 307. is based on the US clotui prices in 1974-5. 16. If small-scale enterprises are used to construct these infrastructure 168 Notes and References components, however .. - cost may be reduced at the same time thz extra employment is generated. 17. Moreover, most of the small-scale industries examined in the Sierr Leone study did not use elaborate buildings but only simple workshop generally located close to their principal markets. 18. The intermediate linkages are often much greater, however, in othe countries with higher levels of industrialisation. Small firms subcon tracting to large firms are important inJapan and other Asian countries Selective Bibliography 19. See Chapter 7. 20. Derek Byerlee, J. Tommy and H. Fatoo, Rural-urbanmigration. 21. Government of Sierra Leone, 1974, NationalDevelopment Planp. 186 22. Government of Sierra Leone, An industrialextension programmefo Allal, M., Chuta, E., Cottage industries and handicrafts: Some guidelines the Republic of Sierra Leone (Freetown, Ministry of Trade and Indus for employment promotion (Geneva, ILO, 1982). try, 1982). Anderson, Dennis, Khambata, Farida, Small Enterprise 23. See Government and Development of Sierra Leone, 'A bill entitled the Small-Scale Indus Policy in the Philippines:A Case Study, World Bank Staff Working Paper tries Act, 1981' in Supplement to the Sierra Leone Gazette Vol. CXI #468 (Washington, DC, 1981). No. 38 (Freetown, Government Printing Department, 16 July, 1981). Austin, V., 'Approaches to rural development: Lessons of a pilot project in Nigeria', in InternationalLabour Review (Geneva, ILO) Vol. 114 No. 2, July-Aug. 1976. __ Rural industrial development: A practical handbook for planners, project managers andfield staff (London, Cassell, 1981). Bank of Sierra Leone: 'Notes and comments', in Economic Review (Free town) Vol. 22 Mar. 1969. __ Annual Report (Freetown, 1975). Bhalla, A.S. (ed.), Technology and employment in industry (Geneva, ILO, 2nd ed., 1981). Bottomley, A.: Factorpricingandeconomic growth in under-developed rural areas (London, Crosby Lockwood, 1971). Brandt, H., Schubert, B., Gerken, E., The industrial town as factor of economic and social development: An example ofJingalUganda(Munich, Weltforum Verlag, 1972). Byerlee, D., Tommy, J., Fatoo, H., Rural-urbanmigration in Sierra Leone: Determinantsand policy implications African Rural Economy Paper No. 13 (East Lar.sing, Michigan State University, 1976). Byerlee, D., Eicher, C.K., Liedholm, C., Spencer, D.S.C., Ruralemployment in tropicalAfrica: Summary of findings African Rural Economy Working Paper No. 20 (East Lansing, Michigan State University, Feb. 1977). -'Employment-output conflicts, factor price distortions and choice of technique: Empirical results from Sierra Leone', in Economic Develop ment and Cultural Change (Chicago, Chicago University Press) Vol. 31 No. 2 Jan. 1983, pp. 315-36. Callaway, A., 'From traditional crafts to modem industries', in Lloyd, P.C., Mabognuje, A.M., and Awe, B. (eds), The city of Ibadan:A symposium on its structureand development (Cambridge University Press, 1967). Child, F.C., Kaneda, Hiromistsu, 'Links to the Green Revolution: A study of small-scale agriculturally related industry in the Pakistan Punjab', in Economic Development and CulturalChange (Chicago University Press) Vol. 23 No. 2 Jan. 1975. Chuta, E., A linearprogramminganalysis ofsmall-scale industriesin Sierra 169 170 Selective Bibliography Selective Bibliography 171 Leone, unpublished Ph.D. dissertation (East Lansing, University, 1977). Michigan State Oxford Economic snMciaSte Papers(Oxford, The Clarendpn Press), Vol. 20 No.-2 The economics of Gara (tie-dye) industry itzSierra Leone African Rural Economy Government of India, Small-scale Working Paper No. 26 (East Lansing, Michigan industries.in India: Policies, programmes State and institutionalsupport University, 1978). (New Delhi, Ministry of Industry, Sep. 1980). Government of Sierra Leone: Techniques of production, efficiency and profitability Quarterly Trade Statistics (1963-73) (Free Leone clothing industry in the Sierra town, Central African Rural Economy Programme Working Statistics Office). - 1963 Population Census of Paper No. 30 (East Lansing, Michigan State University, Sierra Leone, Volumes 1-3 (Freetown, Choice of Jan. 1980). Central Office appropriatetechnique in the African bread industry of Statistics, 1965). special reference with __ Customs and excise to SierraLeone (Geneva, ILO, Jan. 1981; mimeographed tariffs (Freetown, Government Printer, 1967). World Employment _ Report on the survey of business and Programme research working paper, restricted). industry, 1966/67 (Freetown, 'Upgrading the managerial Central Statistics Office, i970) process of small entrepreneurs in West Africa', in Public Administration __ Trade and industry bulletin: Republic anniversary and Development (Sussex, John Wiley issue (Freetown, Sons) Vol. 3, 1983. and Government Printer, 1971). Chuta, _ National accounts of Sierra Leone, E., Liedholm, C., The role of small-scale industry 1963/64 to 1971/72 (Freetown, generationand ruraldevelopment: in employment Central Office of Initialresearch resultsfrom SierraLeone, Statistics, 1973). African National development plan, 1974175 Rural Employment Paper No. 11 (East Lansing, to 1978/79 (Freetown, Govern University, 1975). Michigan State ment Printer, 1974). National accounts of Sierra Rural non-farm employment: A review of the Leone, 1975/76 to 1979180 (Freetown, Rural Development state of the art, MSU Central Statistics Paper No. 4 (East Lansing, Michigan State Univer- Office, June 1981). _ An industrial extension sity, 1979). programmefor the Republic of Sierra Leone Chuta, E., Liedholm, (Freetown, C., Roberts, 0.. Tommy, J., Employment growth Ministry of Trade and Industry, 1982). change and Harper, M., Consultancyfor in Sierra Leone small-scale industrv: 1974-1980, small businesses: The concept of trainingcon Economy African Rural sultants (London, Programme Working Paper No. 37 (East Lansing, Intermediate Technology Publications, 1978). State Michigan Harper, M., Soon, T.T.: University, Aug. 1981). Small enterprises in developing countries:Casestudies Chuta, E., Steacy, M.A and conclusions (London, A case studv of Gara technology in Sierra Intermediate Technology Publications, 1979) (Njala, University Leone Ho, S.P.S., The ruralnon-farm College, 1975; mimeographed, unpublished). sector in Taiwan, Studies in Employment and Rural Development, No. 32 (Washington, E. Chuta and S.V. Sethuraman (eds), Rural DC, World Bank, Sep. 1976, ment in Africa Small Industries and Employ- and Asia: A Review of Programmesand Policies mimeographed). (Geneva, -- Small-scaleindustries ILO, 1984). in two rapidlygrowing less-developed countries, Cole, M.J., Korea and Taiwan: A study Hamilton, D., Indigenoustechnolggy in Sierra of their characteristics,competitive bases and Sierra Leone (Freetown, productivity (Washington, Leone, Government/ECA/UN ICEF Report, 1976). DC, IBRW-,, 1978). Dhar, P.N., Lydall, Holtermann, S., Intermediate H.F., The role of small enterprisesin Indian technology in Ghana:Theexperienceofthe Technol development (Bombay, economic ogy ConsultancyCentre Asia Publishing House, 1961). (London, Intermediate Technology Group, 1979). Eckaus, R.S., Appropriate Huddle, D., An analysis of savings technologiesfor developing countries (Washing- behaviour of a group of Colombian ton. artisan entrepreneurs, Program DC, National Academy of Sciences, 1977). of Development Studies, Paper No. 81 Falcon, W.P., (Houston, Rice 'Agriculture and industrial inter-relationships in University, 1977). tan' West Pakis- Hymer, S., Resnick, S., in Journalof Farm Economics (Ames, Iowa State 'A model of an agrarian economy wits non-agrarian University) Vol. 49 activities', in No. 5, 1967, pp. 1139-1157. The American Economic Review (Wisconsin, George Farbman, Co.) Banta M. (ed.), T're pisces studies. 4ssisting the Vol. 59 No. 4 1969 pp. 493-506. smallest economic ILO: Fiscal measures activities of the urban poor (Washington, for employment promotion in developing countries Faroog, J., Faroog, DC, AID, Sep. 1981). G.M., Adepou, A., 'Characteristics and (Geneva, 1973). generation employment Sharingin Development: of small scaie industries in Western Nigeria' A Programmeof Employment, Equity, and Development in The Pakistan Growth for the Review (Karachi, Institute of Development Philippines(Geneva, International Labour Office) Vol. XV Economics), - Ensuring No. 4 1976, pp. 382-405. equitable growth: A strategy for increasing employment, Fowler. equity and basic needs satisfactionin Sierra D.A., 'The Informal Sector in Freetown: Opportunities Leone (Addis Ababa, Jobs and Employment', in for Self- Skills Programme S.V. Sethuraman (ed.), ThelnformalSectorinDeveloping for Africa, 1978). Countries:Emnployment, Iwuji, E.C., Okorafor, A.E., Small-scale Poverty andEnvironment (Geneva, 11 0, 1981). industries, South Eastern and Frank, Jr., C.R., 'Urban Benue PlateauStates of Nigeria unemployment and economic growth in (Nsukka, University of Nigeria, 1975). Africa', Johnson, B.F., Kilby, P., Agriculture and structuraltransformation:Econ 172 Selective Bibliography omic strategies in late developing Selective countries (London, Oxford Bibliography University PgJ. 173 P r e s s , 1 9 7 5 1 .P a g e , J r., J...M . , ' EEooisoc o n o mie s of s cclicma le , inco me dis trib u tio n and sm al l P raees . lenterprise itiuinadsal Johnson, H.G., Economic policies towards less-developed promotion in Ghana's countries (New timber industry', in Food Research York, Praeger Publishers, tute Studies (Stanford, Food Research Insti Karsten, D., The 1968). Institute), Vol. XVI No. 3 1977-78, economics of handicrafts in traditional pp. Weltforum societies (Munich, 159-82. Verlag, 1972). pSmall enterprisesin African developme12: Kilby, P., African A survey World Bank Staff enterprise: The Nigerian bread Working Paper No. 363 (Washington, ford University, industry (California, Stan- DC, World Bank, Oct. 1979). 19(3). Phan-Thuy, N., Betancourt, R.R., Industrialisation Winston, G.C., Kabaj, M., Industrial in an open economy: Nigeria, capacity and employment promotion: 1945-66 (Cambridge, Case studies of Sri Lanka, Nigeria, The University Press. 1969). '-:, occo and an over-allsurvey of other King, K., The African " developing countries (Westmead, Artisan (New York, Teachers College iand, Gower Publishing, King, R.P., Byerlee, Press, 1977). Ranis, G., 1981). D., Income distribution, consumption 'Production functions, market imperfections consumption patterns and opment', and economic devel linkages in ruralSierraLeone in The EconomicJournal(Cambridge No. 16 (East Lansing, African Rural Economy Paper University Press), Vol. 72 Michigan State University, 1977). No. 286 June 1976, pp. 344-354. - ,'Factor Intensities Resnick, S.A., and Locational Linkages of Rural 'The decline of rural industry under Patterns in Sierra Leone', Consumption comparison export expansion: A American Journalof Agricultural among Burma, Philippines and Thailand, Vol. Economics, nal of 1870-1938', in Jour 60, No. 2 1978, pp. 197-206. Econom'c History (Wilmington, Eleutherian Lawson, R., Kwei, E., African Mills Historical Li entrepreneurshipand economicgrowth: brary), Vol. 30 No. 11970, A case Saylor, R.G., pp. 51-73. sudv of thefishing industry of The economic system ofSierra Liedholm, C., Research Ghana(Accra, Ghana University Press, Leone (Durham, Duke Univer on employment in the ruralnon-farm 1974). sity Press, 1967). sectorin Africa Schadler, African Rural Employment Paper K., Crafts,small-scaleindustries University, No. 5 (East Lansing, Michigan and industrialeducation in Tanza 1973). State nia (Munich, Weltforum Schatz, Verlag, 1968). Liedholm, C., Chuta, S.P., Nigerian Capitalism (Berkeley, industries in Sierra E., The economics of rural and University of California Press, Leone African Rural Economy urban small-scale 1977). Lansing, Michigan Paper No. 14 (East Schmitz, State University, 1976). H., 'Growth constraints on small-scale Linsenmeyer, countries: manufacturing in developing D.A., Economic analysis of alternative A critical review', in World Development opment of SierraLeone marinefisheries strategiesfor the devel- (Oxford, Pergamon African Rural Economy Programme Press), Vol. 10 No. 6 1982, pp. 429-450. Working Paper No. 18 (East Lansing, Sethuraman, S.V., The Mabawonku, Michigan State University, 1976). urban informal sector in developing A.F., An economic evaluation ployment, poverty coun!ries: E'm of apprenticeshiptraining and environment (Geneva, ILO, Western Nigeria small-scale it Shetty, M.C., Small-scale 1981). industries African Rural Economy and household industriesin a developing 17 (East Lansing, Michigan State Paper No. (Bombay, economy University, 1979). Asia Publishing House, 1963). Marris, P., Somerset, A., The African Shippers, F., 'Development enterprise:A study ofentrepreneurship and promotion of small industry development in Kenya and Industry Bulletin sector', in Small (New York, Africana Publishing Corporation, for Asia andFarEast (New McBain, N.S., 1972). Singer, York, United Nations, 1970). Pickett. J., 'Footwear production in Ethiopia H.W., 'Small-scale industry in of appropriate - A case study African economic development', technology', in The Journal Robinson, E.A.G. (ed.), Economic Development in (Cambridge University of Modern African Studies for Africa South of the Press), 13 Mar. 1975, pp. 415-427. Sahara (New York, St. Martin's Meredith, G.G., Smith, Press, 1965). Nelson, R.E., Neck, P.A., The Practice V.E. et al., Household food consumption of Entrepreneur- Development in Sierra Leone, Rural ship (Geneva, ILO, 1982). Working Paper No. 7 (East Lansing, Morawetz, D., 'Employment implications Michigan State Univer of industrialisation in developing sity, 1979). countries', in Economic Journal Spencer, D.S.C., (Cambridge University Press) Byerlee, Derek, Technical Change, 1974, pp. 491-542. Vol. 84 FarmerDevelopment: Labor Use, and Small Evidence from SierraLeone, Nafziger, E.W., African capitalism: Working Afican Rural Economy A case study in Nigerian entrepreneur- Paper No. 15 (East Lansing, Michigan ship California, Hoover Institution Spencer, D.S.C., State University, 1976). Press, 1977). May-Parker, I.I., Rose, F., Employment Nag, A., Small industries: Aspects comes in the riceprocessing efficiency and in of their mortality', in The Economic industry of SierraLeone African Times (New Delhi, 6 Oct. 1980). Working Paper Rural Economy No. 15 (East Lansing, Michigan State Neck, P.A., Small Staley, E., Morse, University, 1976). enteprise development: Policies R., Modern small industryfor agement Development Series and programmes Man- developing countries (New No. 14 (Geneva, ILO, 2nd imp., York, McGraw-Hill Book Company, Nihan, G., Demol, E., Jondon, 1979). Steel, W.F., 1969). C., 'The modem informal sector 'Import substitution and excess capacity in Lomd', in Economic in Ghana', in Oxford InternationalLabourReview Papers(Oxford, The Clarendon (Geneva, ILO), Vol. 118 No. 5 Sep.--Oct. Press), Vol. 24 No. 2 July 1972, 1979. pp. 212-40. 174 Selective Bibliography - Small-scale employment and productionin developing dencefrom Ghana (New York, countries: Evi Stevens, Y., Technologies Praeger Publishers, 1977). for rural women's activities: pects in Sierra Leone (Geneva, Problems andpros ILO, Oct. 1981; mimeographed Employment Programme research World Stewart, F., James, working paper; restricted) J., The economics of countries new technology in developing (London, Frances Pinter, University of Ife: Small-scale 1982). industriesin Mid-West (Industrial Kwaraand Lagos States Wahlman, Research Unit, 1978). M., Chuta, E., 'Sierra Justine Leone ic:sist-dyed textiles', M., and Schwarz, Ronald in Cordwell, Index A. (eds), Fabrics of Culture Watanabe, S., 'Reflections (Paris, on current policies for promoting small enter- Agriculture: Vol.prises 110 and subcontracting', in International and intersectoral Wilcock, D.,No.Chuta, 5 Nov. 1974. Labour Review (Geneva, ILO), 22, 23, 24; production E., 'Employment in rural linkges, 55-6, 149; policies, industries in Eastern Upper 144-5 output and value added hours of, by 149; policies, for, categories, industrial opu7 utt and vae ,; asesona riatiosf, July-A ug. 1982, pp. 455-468. 40-i, 42-4; o Wilde, J.C. de, The development in ara,2 7 of African i,vate Boliland Paper (Washington, enterprise, Vol. ic area, (Washington, DC, WorldDC, Feb. 1978). araisi f World Bank, Employment Bank, 1971). ,4 ;Bo, 6. anddevelopment ofsmall learning 18 enterprisesSector Policy fees for, 45, 47 Bread-tzaking, p RraetpieT and emolovmenonfr7 Apprenticeships, 47-8, baPorietehoof, 113-33;13-0; baking n C b "(Washin"•.gton, DC, 139, 140, technology ndt Jan. 1978). 149, of, 113-20; and tonDTrade C ,1982 and emn,- 151; duration of, 45, 46 bakeries,comparative p c es o rindustrialm adevelopment d 128-30;static economicsanalysis of (Washing. Baking:hours annual m ean capital for, 34;for, annual 41, 43; mean capital labour stock of, 123 5;-- and im plications of inputs inputbakery study, 132-3; labour in, 120; and location for, 32, 120-1; employment in, factors, 113, 128; equipment, 113-16; technology and qpality, and factor proportions 127-8; profit rates in, 124, in 125; industry, 123, 124, 125; and Sierra Leone bakeries in African context, 126-7. imported inputs for, See material 27, 142; also Baking inputs for, 27; output Buildings, and value added for, 22, 24; 28, 29, 30, 31 Business failures, 80-1, 151 production processes 116-20; profit in, 58-9, rates in, 67, 69, Capital: 124, 125; and seasonal market, 37, 138; variations in output, 23; opportunity cost of, 66; technical efficiency recovery factor, 29; services of, 124. See and also Bread-making stock inputs, 28-31; stock, Banks, annual commercial, 30, 37, 138, mean, for five industrial categories, 34; 151 stock per Blacksmithing: and agricultural worker, 64; working, linkages, 28. See also Expansiondefined, capital; 55-6; annual mean capital stock for, 31, 34; Initial capital annual mean Carpentry: annual meais labour hours for, stock capital 43; apprenticeship for, 34; annual mean term for, labour 45; capital inputs for, 30, 32, hours for, 41, 42; 148; capital inputs for, 30, 32; demand for output of, 56; excess capacity imported inputs for, in, 39; output 27, 142; imported inputs for, 27, and value added for, 142; 22, 23, 24; production 175 176 Index Index 177 processes in, 58; profit rates in, clothing firms, 92; of gara capital inputs for, 30, 32, 102; output and value added of, 19, 67, 69; reinvestment patterns dyeing firms, 103-9 efficiency of representative 21, 22-3, 24-5 in, 83-4 Electricity, 28, 88 firms in, 103-9; excess capacity Industrial estates, 142, 151 Closure rates. See Entry and Employment: by branch of activity, in, 39; export markets for, 56, Industrial policies, large-scale, closure rates 76-7; generation, 1, 65; growth 110, 152; family labour in, 41, 14,3-4 Clothing industry, 87-95; budget (1974-80), 75-9; by industrial 101; and imported inputs, 27, Infrastructure policies, 142-3 analysis of representative firms categories, 19-20; in large-scale 141, 142; initial capital for, 35; Initial capital, 31-7, 70-1, 73, 138, in, 89-95; economic efficiency industries, 1; in rural small- labour inputs for, 101-2; 149; and profits, 70, 71, 73, of, 92; factor proportions for scale industries, 13, 18, 19-20, labour intensity of, 96; and 138; sources of, 35-7 representative firms in, 90; and 50; sectoral growth patterns marketing of cloth, 109-11, Innovation, 82-3, 150, 152 implications of study, 95; of (1974-80), 75-80 143; material inputs for, 27, Inputs: capital services and stock, overview of, 88-9; and profits Energy costs, 129-30, 131, 132 102-3; and microeconomic 28-31; initial capital, 31-7,70--1, among firms, 92-5; technical Entrepreneur, role of, 70-4 analysis of firms, 103-9; output 73, 138, 149; labour, 40-51; efficiency of, 91. See also Entry and closure rates, 80-2 and value added for, 22, 24; material and services, 27-8 Tailoring Equipment, 28, 29; baking, 113-16; policy implications for, 111-12; Interest rates, 30, 138, 139, 151 Colombia, 17 blacksmithing, 58; carpentry, production inputs for, 101-3; Intersectoral linkages, 55-6, 143-4, Commercial banks, 30, 37, 138, 151 58; clothing, 88 production process in, 58, 149 Co-operatives, 110-11 Excess capacity, 37, 39, 148 97-9; profits in, 67, 69, 106-9; Kenema, 6, 18 Country cloth, 80 Expansion capital, 37, 38, 71, 138; quality and quantity control Kenya; baking industry of, 131; Credit Guarantee System, 139 and profits, 71, 73, 138; problems in, 100-1; technical capital stock-labour ratio of, Credit policies, 137-9, 151 sources of, 38 efficiency in, 103-9; technology 65; and fixed assets of small Experience, and profits, 70, 72, 73 of, 96-101 industries, 31; initial capital Demand, 52-7; foreign market, 56, Exports. See Foreign markets Ghana: z'apital stock per worker in, investments in, 35 110, 150; and income Extension programme, 146 65; fixed assets of small elasticities, 53-5, 149; and industries in, 31 Labour, 40-51, 149; hours, annual intersectoral linkages, 55-6, Factor prices, 3 Goldsmiths: annual mean labour mean, by industrial catogories, 143-4, 141) Factor proportions: and alternative hours for, 43; annual mean 40-1, 42-4; intensity, 3, 64-5, Development )rdinance of 1960, production processes, 57-62: in capital services inputs for, 33; 150; intensity and profits, 69. 140-1 baking industry, 123, 124, 125; output and value added for, 25 See also Apprentices; Family Directorate of Small Tndustries and for representative clothing Gross Domestic Product, 15, 16 labour; Hired labour; Handicrafts, 145 firms, 90; variations in, by firm Proprietors Distribution of industrial firms, size, 64-5; variations in, by Handicrafts, 145. See also Gara Labour-capital ratio, 59, 60, 61-2, 13-15, 18-20 location, 62-3 dyeing 63; and firm location, 62-3; for Family labour, 40, 41, 101; annual Hired labour: annual mean hours, traditional small-scale firms, 63 Economic efficiency. See Efficiency, mean hours of, by industrial by industrial categories, 41, Large-scale industries: and economic categories, 42-4 42-4; opportunity cost of, 66 economic policies, 143-4; Economic policies, 137-45; Farm: income, 2, 55; labour, 152 Hunger season, 80 employment in, 1, 15; growth agricultural, 144-5; credit, Fixed assets, 28, 31 rate of, 1-2; and import duty 137-9, 151; fiscal, 140-2, 152; Flour, 121, 131 Imports: duties on, 141-2; and exemptions, 141; and linkages infrastructure, 142-3; Foreign markets. 56, 150; for gara material inputs, 27 with small-scale industries, 55, large-scale industrial, 143-4; cloth, 56, 110, 152 Import substitution policy, 1, 95 143-4; input-capital ratios of, technical and managerial Formal education. See Education, Income: agricultural, 144; 65; value added for, 15-17 assistance. 139-40 formal distribution of, 78; and Learning fees, 45-7 Economic pronms 'ee Profits Freetown, 6, 127 elasticities of demand, 53-5, Linkages. See Intersectoral linkages Education, formal: and profits, 70, 149; farm, 2, 55 Location: and annual mean capital 72, 73, 139; of proprietors, Gara dyeing, 96-112; annual mean Industrial categories: annual mean services, 30-1, 32-3; 49-50 capital stock for, 31, 34, 148; capital services input by, 30, distribution of small-scale firms Efficiency, economic: of bread- annual mean labour hours for, 32-3; distribution of firms, by by, 13-15; and factor making firms, 124-5; of 42; apprenticeship in, 41, 45; size and location, 18-20; proportions, 62-3 Index 179 178 Index 7 68 6 9 72, 73; and formal education, returns to proprietors,6 , - Technical efficiency: of baking Makeni, 6, 18 124; of clothing firms, inputs in bread 70, 72, 73, 139; of gara dyeing Small-Scale Industries Act 1981, firms, Management: 89, 91; of gara dyeing firms, industry, 121-3; inputs in gara firms, 67, 69, 106-9; and initial 146-7 103. See also capital, 70, 71, 73, 138; and Small-scale industry, defined, 15 103-9 dyeing, Techqical training. See Proprietors record-keeping, 70, 72-3; Supply, 57-67; and alternative 139-40, 151 reinvested, 83-4, 138 production processes, 57-62; Apprenticeships Managerial training, adoption of, in baking gara cloth, 109-11, Proprietors: and age of small-scale and factor proportions, 59-62; Technology: Marketing:143; policies of for gara industry, firms, 51; annual mean labour and variations in factor industry, 126-7, 130-2, 133; hours for, by industrial proportions by firm size, 64-7; -choices in small-scale 111-12 3 for bread- categories, 42-4; and and variations in factor industries, Material inputs, 27-8; Gara dyeing 121; for gara dyeing, apprenticeship system, 47-8; proportions by location, 62-3 Tie-dye. See making, Training, 139-40, 151. See also 102-3 characteristics of, 48-51; Survey methodology: 1974-5, 4-6; formal education of, 49-50; 1980, 6-7 Apprenticeships Mat making, 18 114 18, 80 returns to, 65-9 Synthetic dyestuffs, 97 Tunnel oven, Metal working, Tunisia, 17 Ministry of Trade and Industry, 146 Multiple deck oven, 114 Radio-repairing: annual mean Tailoring: annual mean capital labour hours for, 44; capital stock for, 34; annual mean Urban small-scale industries, 18, National Development Bank, 139 services inputs for, 33; output labour hours for, 41, 42; 19-20,21 NationalDevelopment Plan (1974), and value added for, 25 apprenticeship term in, 45, 46; 2, 28, 145, 147 Record-keeping, 50; in baking building rentals for, 31; capital Nigeria: baking industry in, 130; industry, 123; and profits, 70, services inputs for, 30, 32; Value added, 15-17; output and proprie.-.s in, 49, 50; small- 72-3 employment and value added, 22-7; by small scale industrial employment in, Reel oven, 114 establishments, 18; excess scale industres, 18, 20-1 17 Reinvestment patterns, 83-4 capacity in, 39, 89; and Vehicle repair, 18 Rentals: building, 29, 30, 31; imported inputs, 27, 141, 142; annual mean Occupational mobility, 48 equipment, 29, 35, 95 initial investment for, 35; Watch-repairing: of Rice, 55, 80 material inputs for, 27; output labour hours for, 43; capital Output, 22-7; demand for, 33; output 52-7; Roads, 142-3 and value added for, 18, 22, services inputs for, small-scale firms, added for, 25 variations in, 23, 26-7 Rotary peel oven, 115 23, 24, 88; production and value seasonal 28, 142 Output-capital ratios, 3, 59-60, Rural Employment Research processes in, 58; profits in, 68; Water, 61-3, 65; for modem and Project, 4, 6 reinvestment patterns in, 83-4; Weaving, 18 season, 5 traditional processes, 60; for Rural small-scale industries, 18, and seasonal variations in 'Wet' 148; employment in, 18, 19-20, output, 23; workshops, 88. See Women, 41, 96 traditional small-scale capital, 28, 30; in bread 63 150; value added by, 18, 21 also Clothing industry Working industries, 120-1 Output-labour ratios, 59, 60-1, 63 Tariff rates, 141-2 industry, Oven types, 58-9, 113-15; Seasonal variations in output, 23, distribution of, 115-16, 117 26-7 Self-financing, 37-8, 149 Peel brick oven, 115 Service inputs, 27-8 Philippines, 17 Sewing machines, 88, 95 Production processes: and factor Shoe-repairing: annual mean labour proportions, 57-62; modern hours for, 44; capital services and traditional, 58, 60, 67, 69, inputs for, 33; output and 150 value added for, 25 Profits, 3, 65-9; of baking firms, Small-scale industrial categories: 124, 125; of clothing firms, distribution of, by size and 92-5, and expansion capital, 71, location, 19-20; output and 73, 138; and experience, 70, value added by, 21, 24-5; and