(PDF) Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Main Features of

Micro and Small Manufacturing

Enterprises in Ethiopia

Baseline Survey Report

Mulu Gebreeyesus

Abebe Ambachew

Tigabu Getahun

Berihu Assefa

Girum Abebe

Seid Hassen

Haileselassie Medhin

ETHIOPIAN DEVELOPMENT RESEARCH INSTITUTE

(EDRI)

Main Features of Micro and Small

Manufacturing Enterprises in Ethiopia

Baseline Survey Report

Mulu Gebreeyesus1

Abebe Ambachew

Tigabu Getahun

Berihu Assefa

Girum Abebe

Seid Hassen

Haileselassie Medhin

1 For any query, the corresponding can be contacted on mulu.yesus@gmail.com

Ethiopian Development Research Institute (EDRI)

Entrepreneurship and Small Business

Development (ESBD) Research Programme

Printed in Addis Ababa

ISBN: 978-99944-72-63-5

Cover Photo:- by the EDRI Survey team

i

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

About the Authors

Mulu Gebreeyesus (PhD): is a senior researcher at the Ethiopian

Development Research Institute (EDRI) and coordinator of entrepreneurship

and small business development (ESBD) research program. Between

2008 and 2014, he had been working as a senior researcher at the United

Nations University (UNU-MERIT), Maastricht the Netherlands. Prior to

this, he served as a Research Fellow and Head of Poverty and Sectoral

Development Directorate at EDRI for the period 2006-2008. Dr. Mulu has

published widely in the eld of development economics and particularly

competition and enterprises dynamics (investment, growth, entry-exit

and productivity), the practice and impact of industrial policy, innovation

and small businesses dynamics in developing countries. He holds a PhD in

Economics from the Department of Economics, University of Gothenburg,

Sweden in 2006.

Abebe Ambachew (MSc): is a researcher associate at the Ethiopian

Development Research Institute (EDRI). He obtained his BA degree

in Economics from Haramaya University and MSc in Economics from

Addis Ababa University. He taught economics for more than a decade at

Haramaya University and has good research experience and publications

records on various issues. He is currently engaged in dierent research

projects, with other fellows in EDRI, in the area of Micro, Small, and Medium

Enterprise (MSMEs) development. His main research interest includes rm

performance, enterprise development, and economic policy evaluation.

Tigabu Degu Getahun (PhD): is a Research Fellow at the Ethiopian

Development Research Institute (EDRI). He is also a senior researcher at

the Center for Development Research, University of Bonn starting from

2015. His main research interest includes gender, individual and household

welfare, rm dynamics, labor, innovation, inter-sectoral linkages and green

industrialization. Getahun has a long track record of quantitative and

qualitative studies. He obtained PhD in development economics from

university of Bonn, Germany and M.sc in economics from University of

Copenhagen, Denmark and B.A degree in economics from Addis Ababa

University.

ii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Berihu Assefa (PhD): is a Research Fellow at the Ethiopian Development

Research Institute (EDRI). His research undertakings mainly include industrial

development: learning technology and better management skills from

abroad, Micro and Small-scale Enterprises development, the implications

of East Asian development for Africa, and trade policy analysis.He received

his Masters and PhD Degrees in International Development Studies and

Development Economics from the National Graduate Institute for Policy

Studies (GRIPS), Tokyo in 2009 and 2013 respectively.

Girum Abebe (PhD): is a Research Fellow and the Director for Micro and

Sectoral Research Center at the Ethiopian Development Research Institute

(EDRI). He obtained Ph.D in Development Economics from the National

Graduate Institute for Policy Studies (GRIPS) in Tokyo, Japan. He was a visiting

scholar at the University of Oxford, Department of Economics, between

January and March 2017.His main research interests are the applications

of eld experiments and behavioral economics to study the constraints

that face enterprises and job seekers in developing countries.  He has

been actively engaged in eld experiment designs, particularly in the

implementations and impact evaluations of business and management

skills training to young entrepreneurs and the impact evaluation of job

search assistance schemes to young unemployed job-seekers.

Seid Hassen (PhD): is Research Fellow at the Environment and Climate

Research Center (ECRC) in the Ethiopian Development Research Institute

(EDRI). He leads the center’s Sustainable Energy transition research theme.

He holds a PhD in Economics from the University of Gothenburg in 2015.

His research works emphasize on the application of microeconometrics

and experimental economics to energy and agricultural economics. Dr.

Seid has published in journals such as Energy and Resource Economics.

Haileselassie Medhin (PhD): is the Director of the Environment and

Climate Research Center at the Ethiopian Development Research Institute.

He is also a Research Fellow at the Department of Economics at the

University of Gothenburg, Sweden. He earned his PhD in Economics from

the University of Gothenburg in 2013. Dr. Haileselassie has published in

top scientic journals such as the Journal of Environmental Economics and

Management and World Development.

iii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

About the Ethiopian Development

Research Institute (EDRI)

Founded in August 1999 as a semi-autonomous government development

research institute, EDRI’s primary mission is to conduct quality research on

the development of the Ethiopian economy and disseminate the results

to key stakeholders within and outside of Ethiopia. EDRI is sponsored by

the Ethiopian government, ACBF, UNDP, IDRC-TTI and IFPRI/ESSP. For more

information, as well as other publications by EDRI sta and its aliates, see

http://www.edri.org.et

Ethiopian Development Research Institute (EDRI)

P.O.Box 2479

Tel: 251115506068

Fax: 251115505588

Email: info@edri.org.et

Website: http://www.edri.org.et

Addis Ababa, Ethiopia

ABOUT THESE RESEARCH REPORTS

The Ethiopian Development Research Institute (EDRI) Research Reports

contain research materials from EDRI and its partners. They are circulated

to stimulate informed discussions among policy makers, practitioners,

stakeholdersand the public at large. The opinions are those of the authors

and do not necessarily reect that of EDRI’s, their home institutions’ or

supporting organizations’. Comments may be forwarded directly to the

author(s) respective addresses.

iv

Acknowledgements

The International Development Research Center (IDRC) of Canada

and the Norwegian Ministry of Foreign Aairs (through support to the

Environment and Climate Research Center of EDRI) deserve special thanks

for funding the rst year operation of this research programme. Without

their generous nancial support, the collection of this extensive data as

well as the development of this report would not have been realized. We

are also thankful to the European Union (EU) for their nancial support,

made at the right time to disseminate this research output and conduct

further research to come. We hope that the support of these institutions

will continue in the future as well.

We are grateful to the survey team members, particularly, for Samuel

Abera (the programmer and data manager during the survey), the survey

coordinators, eld supervisors, and enumerators for their unreserved eort

to generate high quality data. The respondents to the survey, who gave

us their precious time to answer our questions, with their great ambition

to take part in generating credible information and to collectively tackle

problems faced by MSEs, should be acknowledged. We are also thankful

to dierent government ocials who have supported the survey team to

smoothly conduct the eld work in their respective cities. We beneted

a lot from the thorough edition by Cyndi Berck and we thank her for her

unreserved professional support. Last but not least, we appreciate the

support of the two research assistants, Fiseha Asmare and Fikru Debela.

v

Tables of Contents

About the Authors i

About the Ethiopian Development Research Institute (EDRI) iii

Acknowledgements iv

Tables of Contents v

List of tables ix

List of gures xi

Lists of abbreviations and acronyms xii

Foreword xiv

Executive summary xvi

1. Introduction 1

1.1. Background and Justication of the Research Programme

and the Baseline Survey

1

1.2. Objective of this survey report 5

1.3. Limitations of the study 6

1.4. Organization of the survey report 7

2. Overview on the structure of the Ethiopian economy and MSE

development policies and strategies

8

2.1. Overview on the structure and performance of the

Ethiopian economy

8

2.2. Evolution of MSE policies and strategies in Ethiopia 10

2.2.1 MSE Policies pre-1991 10

2.2.2 The EPRDF Regime period (post-1991) 11

3. Survey methodology: sampling design, data collection, and

processing

16

3.1. Data and sampling 16

3.2. Data quality control mechanisms 21

3.3. Remarks on data handling and management practices of

government oces

22

3.3.1.The Central Statistical Authority (CSA) 22

vi

3.3.2. The Federal Micro and Small Enterprise

Development Agency (FeMSEDA)

23

3.3.3. Regional Micro and Small Enterprise Development

gencies (ReMSEDAs)

24

3.3.4. Trade and Industry Bureaus 27

3.3.5. Federal and regional revenue and customs

authorities

29

3.3.6 Lessons learned and actions required 30

4. Survey Results and Discussion 32

4.1. Basic characteristics of the sample enterprises 32

4.1.1.Size class and geographical distribution of sample

MSEs

32

4.1.2. Distribution of enterprises by sub-sector 34

4.1.3. Firms’ age 36

4.1.4. Ownership type and legal status of sample MSEs 38

4.1.5.Location of sample enterprises 39

4.1.6. Cooperatives: how dierent are cooperatives from

others?

41

4.2. Characteristics of the entrepreneurs 46

4.3. Growth Orientation and Risk Taking Behavior of

Entrepreneurs

49

4.3.1. Entrepreneurial Motivation of the MSEs 50

4.3.2. Occupational choices of eentrepreneurs 54

4.4. Job creation and employment dynamics of MSEs 58

4.4.1. Employment dynamics by type of employees 58

4.4.2. Employment dynamics by type of employees and

size category of MSEs

59

4.4.3. Nature of Employment in MSEs 61

4.4.4.Employment dynamics by gender of workers 61

4.4.5. Educational level of MSEs’ Workers 62

4.4.6. Labor Turnover 62

vii

4.5. Earnings, Working Conditions and Occupational Safety of

Workers

64

4.5.1. Earning of workers 65

4.5.2. Occupational Safety and Health (OSH) of workers 67

4.6. Marketing and Supply Linkages of MSEs 71

4.6.1. Type and Number of Customers (Sales Outlets) 71

4.6.2. Main Sources of Productive Inputs/ Raw Materials

for MSEs

75

4.6.3.Marketing Strategies for Outputs/ Customers of

MSEs

77

4.7. Competition, Technology and Innovation 79

4.7.1. Source and strategy of competition 79

4.7.2. Technology Level and Innovation Practices and

Types

81

4.8. Investment and sources of nancing 85

4.8.1. Investment in xed assets 85

4.8.2. Access to and sources of nance for investment and

working capital

86

4.8.3. MSEs’ access to formal loans 88

4.8.4. Capacity utilization of MSEs 90

4.8.5. Saving behavior of MSEs 91

4.9. Business Development Support and Business Practices of

Firms

92

4.9.1. Entrepreneurs’ Access to Business Development

Support

92

4.9.2. Gender Specic Targeting and Support Experience

of Female Entrepreneurs

100

4.9.3. Firms’ Business Practices 102

4.10. Greening of Micro and Small enterprises 106

4.10.1.Introduction 106

4.10.2. MSE’s Environmental legal requirements,

compliance and inspection

108

viii

4.10.3. MSE’s Waste management Practices 110

4.10.4. MSE’s Energy Consumption, Eciency, and

Conservation

112

4.10.5.MSE’s Greenhouse gas emission reduction activitie 116

4.11. The business environment and growth constraints of

MSEs

118

5. Summary and Concluding Remarks 121

References 124

Appendix A: Additional results from the survey data

disaggregated by sample Cities

131

ix

List of tables

Table 1: Value added of major sectors, share in GDP in % 8

Table 2: The current denition of MSE in Ethiopia introduced in 2011 14

Table 3: Population and sampling distribution of MSEs 17

Table 4: Major sections in the survey instrument 20

Table 5: Geographical and size distribution of sample MSEs 33

Table 6: Distribution of sample MSEs by sub-sector and city (%) 35

Table 7: Legal form of ownership 38

Table 8: Means of acquiring the current business 39

Table 9: Specic location of MSEs by cities (%) 40

Table 10: Initiator of the cooperative business enterprises 42

Table 11: How well members know each other during establishment (%) 42

Table 12: Average values of some variables for cooperative s and

non-cooperative MSEs

45

Table 13: Demographic characteristics of sample entrepreneurs 47

Table 14: Educational status of entrepreneurs 48

Table 15: Motivation for start ups (%) 51

Table 16: Entrepreneurs’ vision to stay in business 53

Table 17: Self-employment preference of entrepreneurs 55

Table 18: Engagement in other businesses 57

Table 19: Employment dynamics by type of workers over time

(for our sample MSEs)

59

Table 20: Total employment distribution by rm size and location in 2016/17 60

Table 21: Employment dynamics by nature of employment over time 61

Table 22: Employment dynamics by gender over time 61

Table 23: Educational Background of MSE workers over time 62

Table 24: Average employees’ turnover by location 62

Table 25: Most Common Reasons for Workers to Quit (%) 63

Table 26: The Most Common Reasons for Workers to be Fired 64

Table 27: Average Monthly Earning of Production Workers (in ETB) 65

Table 28: Average Monthly Earning of Production Workers

Disaggregated by Gender and Cities

66

Table 29: Occupational Safety and Health indicators disaggregated by

the location of rms

69

Table 30: Occupation and safety status of workers by size of enterprises 70

Table 31: Sales outlets, customer base and sub-contracting 73

x

Table 32: Main sources of productive inputs for the MSEs 77

Table 33: Marketing strategies of the MSEs 78

Table 34: Primary Competitiveness Strategy of Enterprises 81

Table 35: Self-Evaluated Level of Technology Level, Disaggregated by

Size and Location

82

Table 36: Innovation, Process and Product Innovation disaggregated by

Location and size

83

Table 37: Motivation to Innovate (%) 84

Table 38: Type and Intensity of Innovation Activities Disaggregated by

Location and Size

85

Table 39: Enterprise’s average investments in various xed assets over

time by location (in Birr)

86

Table 40: Main sources of nance for investment and working capital

by size class

87

Table 41: Formal loan application and success rate by size and location 88

Table 42: Main reason for not applying for a formal loan 89

Table 43: Main reasons for why formal loan application was not successful 89

Table 44: Capacity utilization rate of enterprises by location and size 90

Table 45: Main reasons for capacity underutilization 91

Table 46: Saving behavior of enterprises by size and location 91

Table 47: Access to business development support 93

Table 48: Duration, payment modality, and training providers for MSEs 94

Table 49: Access to business extension and input marketing support services 99

Table 50: Gender specic targeting experience of female entrepreneurs 101

Table 51: Business Practice of MSEs 105

Table 52: MSEs experience on environmental legal requirements 109

Table 53: Average number of times MSEs were inspected on

environmental issues

110

Table 54: Wastewater treatment by enterprise size 111

Table 55: Solid waste storage facilities and management 112

Table 56: Most frequently used energy types by enterprises 113

Table 57: Frequency of electricity outage experienced by the enterprises in a

month, over the last one year

114

Table 58: Enterprise habits of turning o lights and disconnecting

machinery when not in use

115

Table 59: Emphasis given by enterprise to energy eciency when

purchasing new machines/equipment

116

Table 60: Enterprise GHG emissions activities 117

Table 61: Major challenges for the operation and growth of MSEs 119

xi

List of gures

Figure 1: Location of sample cities 19

Figure 2: Sub-sector distribution of sample MSEs by size of

enterprises

36

Figure 3: Age category of sample MSEs since establishment 37

Figure 4: Age category of sample enterprises disaggregated by cities 38

Figure 5: Reasons for MSEs to choose the current location 40

Figure 6: Methods of acquiring the current business premises 41

Figure 7: Average size of cooperative members, by gender and period 43

Figure 8: Primary benet of being cooperative 44

Figure 9: Primary constraints of cooperative form of business 45

Figure 10: Education status of entrepreneurs versus rm size 48

Figure 11: Previous occupation of entrepreneurs 49

Figure 12: Total amount of employment created by population of

enterprises in sample cities by size and time

60

Figure 13: Average monthly earnings of Workers in Birr Disaggregated

by Size of the Enterprise

67

Figure 14: Main source of competition for enterprises in Addis Ababa

and outside

80

xii

Lists of abbreviations and acronyms

AA Addis Ababa

ADLI Agricultural Development Led Industrialization

BA Bachelor of Arts

BDS Business Development Services

BEd Bachelor of Education

BSc Bachelor of Sciences

CAPI Computer-Assisted Personal Interviewing

CO2 Carbon Dioxide

CRGE Climate-Resilient Green Economy

CSA Central Statistical Agency

DK Don’t Know

EC Ethiopian Calendar

ECRC Environment and Climate Research Center

EDRI Ethiopian Development Research Institute

EEA Ethiopian Economic Association

EIA Environmental Impact Analysis

EPRDF Ethiopian People’s Revolutionary Democratic Front

ESBD Entrepreneurship and Small Business Development

ETB Ethiopian Birr

FDRE Federal Democratic Republic of Ethiopia

FeMSEDA Federal Micro and Small-Scale Enterprise Development

Agency

FeSMMIDA Federal Small and Medium Manufacturing Industry

Development Agency

FUJCFoSA Federal Urban Job Creation and Food Security Agency

GDP Gross Domestic Product

GHG Greenhouse Gases

GTP Growth and Transformation Plan

HASIDA Handicrafts and Small Scale Industries Development Agency

ICT Information Communication Technology

IDRC International Development Research Center

IEA International Energy Agency

IGE Imperial Government of Ethiopia

xiii

ILO International Labour Organization

MFI Micro Finance Institution

Mgmt. Management

MoFED Ministry of Finance and Economic Development

MOLSA Minister of Labor and Social Aairs

MoT Ministry of Trade

MSc Masters of Science

MSEs Micro and Small Enterprises

MUDC Minster of Urban Development and Construction

NA Not Applicable

NBE National Bank of Ethiopia

NEPS National Employment Policy and Strategy

NGO Non-Governmental Organization

NPC National Plan Commission

OA Outside of Addis Ababa

OECD Organisation for Economic Co-Operation and Development

OSH Occupational Safety and Health

PASDEP Plan of Action for Sustainable Development and Eradication of

Poverty

PDRE People’s Democratic Republic of Ethiopia

PhD Doctor of Philosophy

PMGE Provisional Military Government Of Ethiopia

ReMSEDAs Regional Micro and Small-Scale Enterprises Development

Agencies

SACCO Saving and Credit Cooperatives

SDPRP Sustainable Development and Poverty Reduction Programme

SOEs State-Owned Enterprises

SSA Sub-Saharan Africa

TGE Transitional Government of Ethiopia

TIN Taxpayer Identication Number

TVET Technical and Vocational Education and Training

UNDP United Nations Development Programme

USD United States of America Dollar

WEDP Women Entrepreneurs Development Programme

WHO World Health Organization

Yrs Years

xiv

xv

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Foreword

The Ethiopian economy has seen remarkable growth over the last decade.

In varying degrees all sectors contributed to the registered economic

growth. Ethiopia has also made a big push in investment in infrastructure

and human development. Cumulatively, this has contributed to increased

employment, income, and poverty reduction. Regarding the structure

of economy, smallholder agriculture and services are dominant sectors.

The structural transformation of the economy through industrialization

as envisaged in the industrial policy is yet to be realized. The current

contribution of the industry sector to GDP is not more than 17%. Particularly,

the current performance of the manufacturing sub-sector is much lower

than expected (with 4% GDP share) and the government has given special

emphasis for this sub-sector to gradually take the lion’s share in country’s

industrial sector.

As can be observed, this is key pillar of the growth and transformation plan

(GTP). One of the key strategy to achieve these targets and build an industry-

led economy is through promoting the development of Micro and Small

Enterprises (MSEs). The Government of Ethiopia has been committed to the

development of the MSE sector as it is stipulated in the MSE development

strategy and other sectors’ policies and strategies. Similarly, in the second

GTP, MSE sector (specially the manufacturing MSEs) is given due emphasis

in view of its strategic importance in generating employment as well as a

learning platform for entrepreneurship and emergency of value adding,

private sector that is willing and able to invest in manufacturing industries.

However, the timely revision of the sector’s policy and strategy as well

as its implementation process should be guided by a systematically

generated, credible, and timely information. The sector suers from lack of

such information and I found this book as an excellent contribution to the

development of the MSE sector produced at the right time to partly ll this

gap. This book is prepared based on the survey data generated through

deliberately designed original research to follow up the state of the micro

xvi

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

and small manufacturing enterprises in urban Ethiopia. It is based on

rigorous analysis by experienced researchers in the eld on scientically

generated fresh data from 8174 micro and small manufacturing

enterprises in 10 largest cities of the country. The book covers a

wide ranges of information on dierent aspects of the MSE development

including on characteristics of enterprises and entrepreneurs; job creation,

sales, marketing, investment, nancing, competition and innovation

activities or performances of MSES; the state of MSE workers and working

conditions such as earnings, labor turnover, and occupational safety;

greening MSEs which clearly shows the interaction of the MSEs with

the physical environment; and the business environments as well as the

major obstacles of the MSEs in dierent geographical areas. I enjoyed the

interesting ndings of the study under all these topics which are worth

studying in such detail and I am sure readers will benet a lot from reading

this book.

Mekonnen Manyazewal

With the Rank of Minster

Executive Director of EDRI

Addis Ababa, 2018

xvii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Executive summary

The Entrepreneurship and Small Business Development (ESBD) research

programme was started in July 2016 with the objectives of producing

evidence-based knowledge through rigorous research and building a strong

and integrated knowledge support system to underpin future policy analysis

in small business development in Ethiopia, with a focus on micro and

small rms in the manufacturing sector. In view of these broad objectives,

between December 2016 and May 2017 a baseline survey was conducted of

manufacturing MSEs operating in the 10 largest cities (Addis Ababa, Adama,

Jimma, Bahir Dar, Gondar, Dessie, Dire Dawa, Jigjiga, Mekelle, and Hawassa),

located in 7 regional states of the Ethiopia.

The survey adopted the national denition and dened microenterprises as

establishments with up to 5 employees and small rms with employment of

6 to 30. Having constructed an exhaustive sampling frame, a census of small

enterprises and a random sample of micro enterprises were undertaken. Our

nal database constitutes detailed information on 8174 enterprises, out of

which 3310 (40.5%) are microenterprises, 4553 (55.7%) small enterprises and

311 (3.8%) enterprises in the medium-size category that employ more than

30 workers. Using face-to-face interviews with MSE owners/managers, we

collected quantitative data with 14 modules, which covered a range of issues

pertaining to MSEs. Follow up surveys will be conducted at two year-intervals

and a longitudinal dataset will be created, aiming at better understanding the

dynamics of MSE development in Ethiopia. This report presents major ndings

on MSE development issues based on the baseline survey, which we believed

will inform and engage policy makers, practitioners, researchers and other

stakeholders. This section tries to highlight the main ndings by disaggregating

the information into sub-sections.

Firms’ and entrepreneurs’ attributes

The survey shows uneven geographical distribution of the manufacturing MSEs

in the 10 largest cities of Ethiopia, where Addis Ababa has the lion’s share. For

xviii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

example, when looking at the small MSEs based on the census, Addis Ababa

alone accounted for about 67% of the 4553 small enterprises in operation in

2016/17. In terms of sectoral distribution, three sectors – namely furniture and

wood working, food and beverage, and metal working workshops – are the

most important sub-sectors, accounting for about 63.72% of manufacturing

MSEs in the selected major cities in Ethiopia.

The MSE sector is dominated by young rms, of which about 60% of sample

enterprises have been in business for no more than ve years. About 95.7%

reported they had established the business by themselves from scratch. This

suggests that entrepreneurship, particularly in the manufacturing sector, is

largely a recent phenomenon in Ethiopia, requiring a conscious strategy to

promote start-ups.

With regard to enterprise location, about 57% operate outside their home in

commercial and non-commercial areas. About one-fth of the sample MSEs

operate in industrial zones, with a large dierence between Addis Ababa

(28.7%) and regional cities (9.8%). While about 40% of the current working

premises of sample MSEs are rented from the private sector, the government

plays a critical role in supplying this facility through rental and leasing, for

about a third of MSEs.

The majority (57%) of the sample enterprises operate in the form of sole

proprietorship, followed by cooperative (14.5%). The survey showed that

cooperatives are more likely than non-cooperative enterprises to get

government support, commonly through the provision of work premises and

skill training, which is considered as one advantage of being cooperative.

However, 38% of cooperatives reported that the diverse interests of members

was the main disadvantage of being a cooperative, leading to the risk of

business termination or withdrawal of members. When comparing the current

cooperative members with members during the initial period, their number

declined by 29.5%. Generally, cooperative MSEs were found to be younger,

driven more by necessity than opportunity, have slower employment growth

rates, and less likely to hire permanent employees as compared to other forms

of business.

xix

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The owner-managers of sample manufacturing MSEs are mostly adults, with

the average age of about 38 years, male, married, and Orthodox Christian.

Young adults (aged 18-35 years) have a share of 48% in terms of owning

and managing enterprises, which is lower than the urban youth population

distribution. The gender gap in owning/managing the MSEs looks substantial.

About 81% of the MSEs were owned and managed by male, but less than a

fth by females.

With regard to human capital of the entrepreneurs, more than 97% have

some formal education, out of whom about 33% completed high school, 17%

TVET (Technical and Vocational Education and Training), and 12% rst degree

and above. The level of TVET graduates (17%) is very low in the context of

the government eorts to massively produce TVET graduates in the hope of

generating a large pool of entrepreneurs.

Growth orientation and risk-taking behavior

Micro and small enterprise operators have diverse motivations and ambitions

for setting up their businesses. As frequently pointed out in the literature, small

enterprise formation is usually considered as a last resort to make ends meet

and support one’s livelihood. This is often true of the informal economy but less

clear for enterprises in the formal sector. Using our sample data, we measure

such attributes in the formal economy by considering entrepreneurs’ motivation

for setting up the businesses. We nd that more than a third of entrepreneurs

initiated their businesses seeking market opportunities. Further, about 38%

of entrepreneurs considered “being own boss” as the main motivation that

underlies their decision to set up their businesses. Unlike entrepreneurs who

roam the informal sector, our results indicate that entrepreneurship was a

conscious choice made by the vast majority of entrepreneurs in the formal

manufacturing sector, seeking better protability opportunities.

Consistent with the opportunity-seeking narrative, we also nd that more than

55% of entrepreneurs strongly prefer running their own business to joining

wage employment. Entrepreneurs also would want a very large monthly wage

premium, ranging from 3806 Birr on average in regional cities to 13,496 Birr in

xx

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Addis Ababa (sample average is 9126 birr) to join the wage sector if an oer

were made. In fact, in the entire sample, only 6% of entrepreneurs are actively

looking for a wage job and only 5% stated that they have a plan to cease

operation in the next ve years. Not surprisingly, about 88% of the entrepreneurs

have expansion plans and 90% think that their businesses have good growth

potential. We also show that about 83% of our sample entrepreneurs operate

a single businesses, 15% have two businesses and 2 % have three business.

This indicates that lateral expansion is not very common among our sample of

enterprises. These results taken together depict a positive picture of the small

enterprise economy in Ethiopia. However, without alleviating the challenges

faced by MSEs, it will be dicult to realize the growth ambitions of enterprises.

Job creation

The total sum of employment generated by the 8174 MSEs in our sample in

2016/17 was 72,584. However, the microenterprises in our database are based

on a sample (only 17% of total population), while the small size enterprises

on evaluated on a census basis. When we estimated the population of the

microenterprises, we nd that about 64,706 jobs were generated by the

microenterprises in the 10 major cities. This increases the total employment

created by the population of manufacturing MSES operating in the 10 largest

cities to 113,705, out of which microenterprises account for about 57%.

The report also shows that most of the jobs created by the MSE sector are

temporary positions – for example, about 77% in 2016/17. However, although

the proportion of permanent employment is small relative to temporary

positions, there has been improvement in terms of a shift towards permanent

jobs, as is shown by the increase in the share of permanent employment from

17 percent in the initial year to 22.7 percent in 2016/17. Similarly, only about

16 percent of MSE rms have a written contract with their workers. These kinds

of terms of employment will have implications for the quality of jobs created

by MSEs.

The manufacturing MSEs predominantly employ male workers. For example,

in 2016/17 nearly 80% of employees were male. This shows that the low

xxi

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

level of participation of women concerns not only entrepreneurship but also

employment and calls for the attention of policy makers and practitioners in

the gender eld. In terms of human capital accumulation of workers, while

overall human capital accumulation has shown slight improvement over

time, it is quite apparent that most of the workers (more than 87 percent) in the

MSE sector are high school graduates and below. TVET and college graduates

constitute only 7 and 6 percent, respectively, in the same period.

Labor turnover, Earnings and Occupational Safety

The report also shows a high turnover of workers. About 3.55 employees left

the average rm in the last 12 months prior to the survey, while 0.48 were red

due to various reasons in the same period. The rm-level average number of

hired workers in the same period was only 4.8, making the net employment

addition only 0.27 per rm per annum. According to the respondent rms, the

most common (81%) reason for workers to quit their job of their own will is to

look for a better salary.

The report indicates that on average MSE production workers receive a basic

after-tax salary of ETB 2069 ETB (USD96) per month in 2016/17. However, since

we do not have data on the number of dependents of each worker, we cannot

comment on whether the sample MSEs oer sucient earnings to permit the

workers themselves and their dependents a level of consumption above the

poverty line. But, as shown above, low pay is the major reason for high labor

turnover, suggesting that the salary level is not sucient to cover the rising

living costs.

The gender-disaggregated analysis unveils a statistically signicant earning

dierence between female and male production workers. The average

monthly earning of male production workers is 20% to 25% higher than their

female counterparts, depending on the year. We are not able to distinguish

to what extent a skill gap between female and male production workers may

explain this dierence. But, when taken together with the low level of women’s

participation as entrepreneurs and employees, this requires the attention of

policy makers in devising instruments to reduce gender inequality.

xxii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Regarding working conditions at the factory, our data shows that there had

been an average of 1.6 work-related accidents in the surveyed enterprises

within 6 months prior to the survey month. Our survey data also provides

evidence that about a third of the sample enterprises have not made their

workers aware of the general occupational safety and health issues in the

factory. This is harmful for both workers and enterprises and hence calls for

policy action.

Marketing and supply linkages

The survey shows that MSEs in the manufacturing sector supply their main

product to a large number of customers and are less likely to depend exclusively

on one or few customers. We observed that more than 50% of the MSEs have

over twenty customers for their main product, while only about 23% of the MSEs

have ve or fewer customers for such a product. In the absence of exclusive

customers, the MSEs are obliged to deal with a large number of customers.

Sub-contracting practices are limited, as only about 6% of the MSEs are

currently subcontracted to other rms. Among those MSEs which are currently

subcontracted to other rms, the majority (about 87%) are subcontracted to

domestic rms, while 25% are subcontracted to the government. Only 3.5%

of such rms are currently subcontracted to foreign rms. Such an absence of

sub-contracting arrangements limits the scope of learning of the MSEs.

The report also presents the sources of productive inputs for the MSEs.

The majority (about 87%) have domestic private enterprises as their main

sources of productive inputs. Only about 9% and 4% of the MSEs have state

enterprises (SOEs) and other sources (e.g., FDIs, non-commercial entities and

direct imports), respectively, as their sources of productive inputs. With regard

to methods of marketing, the majority of the MSEs (89%) rely on their clients,

suppliers, or relatives/ friends as their main avenues of disseminating marketing

information. In fact, 57% of the entrepreneurs stated that clients are their main

channels of marketing information. Other methods of distributing marketing

information, such as other producers, advertisement in the media, business

associations and trade fairs, serve a relatively small (about 10%) proportion of

the businesses as the main channels of such information.

xxiii

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Competition, technology and innovation

Our survey data discloses that the main source of competition for the

manufacturing MSEs currently are formal domestic manufacturing rms. Less

than a tenth of sampled enterprises in the ten study cities identied imports

and informal domestic rms as their main competitor. Providing quality

products and oering competitive prices are the rst and second major

means of competition by the MSEs. On the other hand, forward contracts,

advertisement and credit sales are sources of competition for very few MSEs.

The data also shows that about three-fth of the sampled enterprises had

made an important improvement or change to their business within two

years prior to the survey year. About a quarter of the sampled enterprises had

introduced a new production process and/or new product group. For the vast

majority of these rms, the increasing competition from the formal domestic

sector is the primary motive to introduce new production processes and

product groups. Our disaggregation of the innovation data by the location

of enterprises uncovers that signicantly more MSEs operating in the capital

city made important changes in their businesses and engaged in process and

product innovation, compared with MSEs operating in the regional cities.

Increasing the quality of products, installing new machinery, increasing the

variety of products, and introducing better product design were the dominant

innovation practices reported by enterprises.

Investment and nance

Some of the key observations from the descriptive report on investment

include: rst, average rm investment in xed assets declined from Birr 238,246

in 2014/15 to Birr 152,876 in 2016/17. This is so because major investments in

xed assets usually happen initially – xed investments that happen after the

initial year are usually on expansion projects, which require smaller amounts

than the initial investments in xed assets. Second, an enterprise located in

Addis Ababa invests more in total xed assets than an enterprise located in

other major urban areas in Ethiopia. Third, while enterprises located in Addis

xxiv

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Ababa tend to invest more in land, buildings and other capital goods (such as

factory shades), enterprises located in other major urban areas tend to invest

more in machinery and equipment. This could be due to the fact that it is

relatively easier for rms located in regional cities to access land and working

space at a lower cost than their counterparts in Addis Ababa. On the contrary,

access to land and working space for enterprises located in Addis Ababa is quite

limited simply because the demand is so high and the supply is quite scarce.

The descriptive analysis shows that MSEs’ credit access is very much limited.

Above 70% MSEs had no access to credit from any of the potential external

sources; thus, they rely on their own funds. Of the enterprises that had access to

loans, two-third reported having loan access from MFIs, while only about 15%

from commercial banks. About 40% of the medium-sized enterprises had access

to investment loans, while only 22.5% of micro rms had similar access. This

suggests that formal banking has not been accessible by micro and small rms.

About 61% of enterprises have never applied for formal loans, mostly because

of ‘lack of collateral’. About 46 percent of the rejected loan applicants reported

that their formal loan application was rejected because they could not provide

the required collateral and 20% mentioned the absence of a guarantor (or

joint liability group). This shows that high collateralization is not only a barrier

to applying for credit but also a major reason for rejection of loan applications.

Capacity utilization

Capacity utilization measures the extent to which the productive installed

capacity of an enterprise is being used. The survey results show that capacity

utilization is generally low. The average capacity utilization rate across all size

categories and locations is 54.5 percent. However, as an enterprise gets larger

in terms of size, its capacity utilization seems to improve. For example, while

an enterprise in the Micro size category has an average capacity utilization of

49 percent, it is 63 percent for a rm in the medium size category. Firm that

are located in Addis Ababa have better capacity utilization rates than rms

located in other cities. When rms were asked to identify the key constraints

leading to their capacity underutilization, lack of working capital (57%) and

insucient demand for their products (21%) stood out as the main reasons.

xxv

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Access to Business Development Services

Out of the total of 8174 entrepreneurs in our sample, 3847 (47%) have attended

some sort of training. A larger proportion of entrepreneurs (43%) have

received training after establishing their businesses than before establishment

(23.4%). Entrepreneurs in Addis Ababa and those with larger rms are more

likely to receive training than small rms or those located in the other 9 cities.

Government agencies and TVETs are the main training providers.

Among the dierent types of training that entrepreneurs received after

establishing their businesses, technical skill training (74%) and Kaizen training

(62%) were the more common ones. Marketing management training

and entrepreneurship training were less common. Surprisingly, only a few

enterprises (about 12%) have received any business extension or counseling

services. Moreover, we found that only about 11.3% of the female entrepreneurs

running MSEs in the manufacturing sector have received female-targeted

support services. These support services have been mainly delivered by the

Women Entrepreneurs Development Programme (WEDP), a joint programme

between the World Bank and the Ethiopian government.

Business Practices

In terms of business practices, we found that a large proportion of the

businesses (about 70%) do visit their competitors in order to observe

what prices and products the latter charge and oer. Customers’ demand

assessment has been practiced among the majority (74%) of the MSEs within

the sector. The majority of the MSEs (more than 80% of them) have also

made price and quality comparisons between their own suppliers and the

suppliers of their competitors. Moreover, we nd that 44% of the businesses

set sales targets for the next scal year. Similarly, about 40% of the enterprises

do make comparisons on sales achieved against the targets set on at least a

monthly basis. In addition, we found that more than half of the MSEs in the

manufacturing sector have annual prot and loss statements, while more

than 45% of these enterprises have a strategic plan for their companies.

xxvi

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Greening of MSEs

The greening aspect of micro and small enterprises has not been given

enough emphasis from policy and research perspectives. This baseline survey

attempted to assess MSEs’ greening awareness and practices in terms of

their knowledge of environmental laws and compliance with the laws, liquid

and solid waste management practices, energy eciency and conservation

measures, and greenhouse gas emission reduction strategies. The survey

shows that 82% of the sample enterprises have not done any environmental

impact assessment. A large proportion of the enterprises are fairly condent

of their knowledge regarding environmental legal requirements, which

implies that knowledge is not an issue. More than 80% of all enterprises do

not have wastewater treatment facilities and 90% of the enterprises surveyed

use electricity as a source of energy for their enterprises activities. 80% of the

enterprises turn o lights when rooms are not in use and unplug machines

when they are not in use. More than 75% of the enterprises use energy ecient

light bulbs. Few enterprises use fossil fuel and traditional energy sources such

as biomass fuels. Most of the enterprises do not participate in GHG reduction

actions. Along with promoting the expansion of enterprises for their economic

contribution, government authorities should give due attention to greening

aspects of such enterprises, through research and policy.

Challenges for Operation and Growth of Business

The MSEs were asked to rate among some of the most common challenges

within the sector to the daily operation and growth of their company. In

order of importance, the lack of adequate working premises, lack of access

to credit, and shortage of power supply have been ranked as the three most

important obstacles that call for the attention of all stakeholders working on

MSE development.

1

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

1. Introduction

1.1. Background and Justication of the Research Programme and the

Baseline Survey

With more than 94.3 million inhabitants, Ethiopia has the second largest

population in sub-Saharan Africa (SSA) (CSA, 2017). Similar to many other

SSA countries, the agriculture and urban informal sectors are respectively

the rst and second largest sources of employment in Ethiopia. More than

80 per cent of the population resides in rural areas, relying on subsistence

agriculture for livelihood. On the other hand, the informal sector, which

is also the mainstay of micro and small enterprises (MSEs), absorbs the

largest proportion of urban employment.

Since the early 2000s, the Ethiopian government has initiated and

implemented successive development plans2 and sectoral strategies

to transform the economy and create more and better jobs. Poverty

reduction through employment generation was the core objective in these

development plans and sector strategies. Sustaining broad-based and pro-

poor growth was envisaged to achieve the poverty reduction objective. In

2009, a National Employment Policy and Strategy (NEPS) document was

prepared to coordinate and guide the cross-sectoral intervention needed

to improve employment and its poverty outcomes in Ethiopia.

The previous and present development strategies give due emphasis

to the role of Micro and Small Enterprises (MSEs) for the creation of

employment opportunities and poverty reduction, particularly in urban

areas. Accordingly, the MSE sector was identied as a priority sector for

multifaceted government support. A full-edged MSE Development

strategy was formulated in 1997 and institutionalized a year later by

the establishment of the Federal Micro and Small-Scale Enterprise

Development Agency (FeMSEDA) and Regional Micro and Small-Scale

These development plans are the Sustainable Development and Poverty Reduction Program (SDPRP) 2002/03-2004/05, the Plan of Action for

Sustainable Development and Eradication of Poverty (PASDEP) 2005/06-2009/10, and the ongoing Growth and Transformation Plans.

2

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Enterprises Development Agencies (ReMSEDAs). The strategy was revised

in 2011 with renewed interest and far more ambitious targets (FDRE, 2011).

The vulnerable segments of the society, particularly women and youth,

have been given priority in the employment generating schemes. The

strategy is packed with various promotion and support packages, including

facilitating access to markets, provision of technical and entrepreneurial

training, and improving access to nance and working premises.

The strategy has been translated into medium-term development plans

and due attention is given to its implementation. For example, in the rst

growth and transformation plan (GTP) (2010/11-2014/15), the country had

a target to create about 3 million jobs through the promotion of MSEs

(MoFED, 2010). At the end of GTP I and eve of GTP II, the government of

Ethiopia reported that more than 7 million permanent and temporary jobs

had been created by MSEs ((NPC, 2015; FeMSEDA, 2016), although some

dispute this number as inaccurate. In order to strengthen the role played

by MSEs, it is also envisaged in GTP II to extend the overall support that the

government is providing to this sector.

The role of MSEs in Ethiopia is anticipated to go beyond generating

employment for the urban unemployed. Rather, the sector is at the center

in paving the way to create an industry-led economy. In line with the vision

of being a leading country in Africa in light manufacturing industries by

2025, the manufacturing MSEs have been given great emphasis. As a result,

the federal implementing agency has been restructured, which resulted

in the replacement of FeMSEDA with two separate federal agencies: the

Federal Small and Medium Manufacturing Industry Development Agency

(FeSMMIDA) and the Federal Urban Job Creation and Food Security Agency

(FUJCFoSA). FeSMMIDA is accountable to the Ministry of Industry and is

mandated to foster small and medium manufacturing enterprises (FDRE,

2016a; 2016b).

Despite the unparalleled attention given to MSEs in the development course

of Ethiopia, there is a huge knowledge gap regarding the performance

and current state of the sector. Except for administrative reports, there is

3

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

no systematic evidence about how much and what type of employment

has been generated in the MSE sector during the last decade’s spectacular

economic growth in the country. The administrative reports, primarily

those released by CSA and FeMSEDA, have a number of limitations and

thus neither of them is dependable as far as the state of MSEs is concerned.

Neither the macro-level administrative reports nor the previous attempts

at micro-level case studies on MSEs have produced sucient information

on issues that are important to academia and policy makers. Most of them

focused on only a few aspects of the sector, with limited geographical

scope, and therefore are dicult to depend on (MUDC, 2013; Rijkers et

al, 2010; Haile and Batra, 2016; Lashitew, 2017). As a result, little is known

about the level, type and nature of employment that MSE generates;

the circumstances in which the small business entrepreneurs are

operating; their overall business practices; and to what extent the massive

entrepreneurial schemes underway have created dynamic entrepreneurs

willing to take risks to grow big and thrive. Further, despite the country’s

eorts to promote green and inclusive growth, no information is available

regarding the occupational safety of MSE workers, or on green operations

or environmental friendliness of manufacturing enterprises. The lack of

research on MSEs in Ethiopia is aggravated by the absence of a reliable and

sector-representative dataset.

Therefore, the Ethiopian Development Research Institute (EDRI) initiated

a long-term research programme – namely, Entrepreneurship and Small

Business Development (ESBD) – which aims to ll the existing gap in

knowledge generation and dissemination regarding small business

development in Ethiopia. The research focuses on MSEs engaged in

manufacturing. This research programme was commenced in July, 2016.

The core objectives of the research programme are:

(i) to conduct rigorous research and produce evidence-based

knowledge and impact evaluation that can be used to improve

the existing policies and practices of the government and all other

relevant actors

4

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

(ii) to build a stronger and more integral knowledge support system

and database within the country to underpin future policy analysis

in small business development.

The research programme identied medium-term detailed activities and

work programmes that cover 2016 to 2019. The availability of reliable data

is crucial in conducting sound research and particularly monitoring. One of

the major planned initial activities of this project was, therefore, collection

of baseline data, taking a sample of 8,000 manufacturing MSEs from the 10

largest cities in Ethiopia (Addis Ababa, Adama, Jimma, Dire Dawa, Jigjiga,

Mekelle, Bahir Dar, Gondar, Dessie, and Hawassa). The programme aims to

form longitudinal data by conducting a follow up survey every 2-3 years.

The baseline survey was conducted between December 2016 and May

2017, and generated detail information on a total of 8174 manufacturing

MSEs. The survey was based on a census of small manufacturing enterprises

(with employment of 6-30) and a random sample of micro enterprises

(with employment of 1-5). In terms of size distribution, the data consists

of 3310 (40.5%) micro enterprises, 4,553 (55.7%) small enterprises, and 311

medium-size enterprises (>30 employment), where current employment

exceeded 30 at the time of our survey. Geographically, Addis Ababa has

the lion’s share of MSEs (55%) and the rest are distributed among the nine

other regional cities.

This survey is dierent from the previous ones in at least the following

aspects.

•By utilizing dierent sources, the survey is comprehensive and

based on a complete sampling frame, which includes MSEs formed

with and without FeMSEDA’s support3.

•The geographical coverage is wide, consisting of the 10 largest

cities, representing ve regional states and two city administrations.

•The size of the sample is adequate to detect the characteristics of

dierent sub-groups.

Many other fragmented studies undertaken on MSEs in Ethiopia are based on the FeMSEDA list only. This causes a clear systematic exclusion of

rms that do not have any attachment with the agency.

5

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

•The thematic areas the survey addresses and the depth of the

information it contains are extensive.

•This study was designed not only to take a snapshot of the data

but also to undertake follow-up surveys and construct rich panel

datasets.

This survey report provides the basic facts and gures on the MSE sector

with brief discussions on the implications. It is, thus, of a descriptive nature.

A further analysis is ongoing based on certain thematic areas, which

is reported separately. The present survey report is designed to provide

insights on the current state and challenges of the manufacturing MSEs

sector in Ethiopia and to have wide appeal to a broad audience that

includes policy makers, practitioners, development partners and academia.

This report is presented with appropriate comparison of results based on

enterprise’s size, location, and other attributes. As far as comparison of

results based on location is concerned, the main body of the report mostly

compares cities based on two categories- Addis Ababa and regional

cities. Interested readers can also refer appendix A for additional results

disaggregated by the ten cities.

1.2. Objective of this survey report

The aim of this research report is to highlight the main ndings of the

baseline survey. More specically, this research report has the following

objectives.

•Understand the basic characteristics of entrepreneurs and their

enterprises

•Identify characteristics and constraints of dynamic entrepreneurs

•Evaluate the job creation role of MSEs and their working conditions,

including occupational safety

•Examine the rms’ access to market and nancial services

•Identify the business development supports to and business

practices of rms

•Understand the MSEs’ experience with energy and water

consumption, disposal of wastes, and other environmental issues

6

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

•Identify the major constraints hindering MSEs’ development in

Ethiopia.

1.3. Limitations of the study

Our data set is representative of the micro and small enterprise population

in the 10 largest sample cities. As noted earlier, we conducted a census

of all small businesses and sampled micro enterprises randomly in all

these cities. There are, however, some limitations that should be noted in

discussing our results. First, our data consists of only formal enterprises

that are registered with relevant authorities at dierent layers of the

government and hence we cannot draw any inferences about the informal

economy and how it operates.

Second, we restricted our population to mainly micro and small enterprises.

Enterprises of self-employment (i.e., with no employed worker) as well as

with above 50 workers are excluded from the survey. About 4% of our

sample consists of enterprises that employ between 31 and 50 workers

and hence could be classied as medium sized. Throughout the analysis,

we refer to survey data drawn from these rms in order to suggest some

comparisons with micro and small rms, but we cannot make valid

inferences about the conditions of medium-sized enterprises because we

did not design our survey to collect representative data on them.

Third, our sample is exclusively drawn from enterprises in the 10 largest

urban areas and hence our results are only applicable to the urban

context. Further data is required to draw inferences on rural enterprises

and enterprises operating in other smaller cities not covered by our study.

Fourth, some of the behavioral and skill measurements are based on

hypothetical games and hence may not fully reect behaviors and choices

in real world settings; the usual limitations of non-incentivized behavioral

elicitation experiments apply.

7

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

1.4. Organization of the survey report

The remaining part of the report is organized as follows. The next chapter

provides an overview of Ethiopia’s MSE development policies and strategies

and the sector structure and performance. Chapter 3 presents the survey

methodology, including sampling design, data collection and processing.

This chapter also highlights the features of dierent sources of enterprises’

list sources in the country and the lack of a national database as well as

its implications. Chapter 4 is the main part of this report and presents the

major ndings of the survey results in 11 sub-sections. Finally, the report

closes by presenting summaries and preliminary conclusions.

8

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

2. Overview on the structure of the Ethiopian economy

and MSE development policies and strategies

2.1. Overview on the structure and performance of the

Ethiopian economy

Ethiopia envisages becoming a lower-middle income economy by 2025.

The country has experienced a robust growth rate over the last decade

(2006/07-2015/16), averaging 10.2% per annum (see Figure 2.1). This

growth experience is not only continuous and broad-based, but also much

higher than the regional average, which was estimated to be 5.4% from

2003/04-2014/15 (World Bank, 2017a).

When we see the structure of the Ethiopian economy, it is dominated by

the service and agricultural sectors. As shown in Table 1, the leading role of

agriculture in contributing to GDP has been overtaken by the service sector

since 2010/11. In 2015/16, for instance, the GDP shares of the agriculture,

industry, and service sectors were reported to be 36.7%, 16.7%, and 47.3%,

respectively. Thus, the service sector is becoming the dominant sector,

despite the country’s recent ambition towards industrialization. The share

of the industry sector in GDP has been growing slowly. It rose to 16.7% in

2015/16 from its level of 10.2% in 2009/10.

Table 1: Value added of major sectors, share in GDP in %

Sector

2009/10

2010/11

2011/12

2012/13

2013/14

2014/15

2015/16

2009/10-

2015/16

Average

Agriculture and allied activity 46.1 44.4 43.1 42 40.2 38.7 36.7 41.6

Industry sector 10.2 10.4 11.4 13 13.8 15 16.7 12.9

Manufacturing 3.9 4.0 4.1 4.3 4.6 4.8 - 4.3

Large- & medium-scale4 2.5 2.6 2.7 3.1 3.4 3.7 - 3

Small & cottage industry 1.4 1.4 1.3 1.2 1.2 1.1 - 1.3

Service sector 43.7 45.2 45.9 45.5 46.6 47 47.3 45.9

Source: compiled from NBE (2017) and EEA (2016)

4 This data is based on the CSA’s denition, where large- & medium-scale enterprises are dened as rms engaging 10 or more persons and using

power driven machinery, while those rms engaging fewer persons are considered as small and cottage.

9

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The industry sector is dominated by the construction sub-sector, while the

role played by the manufacturing sector remained stagnant over the last

decade. As shown in Table 1, in the last seven years, the average contribution

of the manufacturing sector to GDP was only 4.3%. Of this, the large- and

medium-scale industries having 10 or more persons engaged had the lion’s

share (on average, 3% of the total GDP), while the role played by small and

cottage industries was limited to 1.3% of the total GDP (Table 1).

However, it should be noted that this information was generated from the

CSA studies based on its own denition of small-scale versus large- and

medium-scale industry. This size class categorization of small or large and

medium is not in line with the ocial denition of the country. CSA denes

small as an establishment with fewer than 10 employees. Therefore, the

data presented in Table 1 regarding the GDP share of small and cottage

industries does not represent the role played by micro and small enterprises

with up to 30 employees, as dened nationally. To our knowledge, there

is no study or national report that has estimated the GDP share of MSEs

based on the national denition as stipulated in FDRE (2011). Taking this

into consideration, we argue that the dierence in GDP share between

large and medium businesses and MSEs may not be as large as presented

in Table 1.

In a nutshell, Ethiopia’s economy exhibited double digit growth in the last

decade or so, but the pace of structural transformation remained slow.

The industry sector is in its infant stage, with a small contribution to the

economy, both in terms of GDP and employment. The role of MSEs is

believed to be signicant, particularly in terms of employment generation

in urban areas. However, the exact gures are not yet known, largely due to

a lack of data that is consistent with the national denition of MSE.

10

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

2.2 Evolution of MSE policies and strategies in

Ethiopia

2.2.1. MSE Policies pre-1991

In the imperial era (ending about 1974), the Ethiopian economic system

was market-led. The then-government introduced some incentives, such

as tax relief, access to land and buildings, and provision of public utilities, in

order to encourage micro and small enterprises (IGE, 1966). However, the

main focus of the industrial policy of the imperial regime was to promote

large-scale manufacturing, which was dominated by foreign nationals. As

a result, the development of MSEs was retarded.

When the Dergue junta took power in 1974/75, the majority of private

sector enterprises (particularly medium- and large-scale) were nationalized.

The role of the private sector was systematically eroded as the government

introduced a single license rule and capital ceiling for investment by

proclamation No.76/1975 (PMGE, 1975). This shrunk the private sector into

micro and small-scale enterprises.

In 1977, the government established the Handicrafts and Small-Scale

Industries Development Agency (HASIDA) by Proclamation No. 124/1977,

with the objective of boosting the economy through the development of

cooperatives in small business activities. HASIDA managed to form many

service and producer cooperatives; however, it did not bring the desired

outcome, as many cooperatives were bankrupt and closed.

The regime was forced to amend its economic policy from command to

a mixed economic system in the last two or three years of the Dergue

period, when the Council of States promulgated two declarations. The rst

one was the Small Scale Industry Development Special Decree No.9/1989,

which allowed the establishment of small-scale enterprises by business

organizations, individual entrepreneurs, and cooperatives (PDRE, 1989). By

replacing the previous legislation (Proclamation No. 76/1975), this decree

raised the capital cap of small enterprises from Birr 500,000 to 2-4 million

11

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

and also allowed the participation of Ethiopians in the Diaspora. The

second was the Special Decree on Investment, No.17/1990, through which

the government declared the total removal of the previous restrictions on

the number of licenses to the private sector, and investors were allowed to

invest in an unlimited number of business activities (PDRE, 1990).

However, these reform programmes were too late to rescue the economic

downturn of that period, as the Dergue regime was toppled in May 1991

and replaced by the Ethiopian People Revolutionary Democratic Front

(EPRDF) regime. In sum, the policy environment of this era was not enabling

for MSEs.

2.2.2. The EPRDF Regime period (post-1991)

i. The early-stage general reform programmes

After overthrowing the Dergue regime, the EPDRF came to power and

formed the transitional government of Ethiopia in 1991. In 1992, Ethiopia

embarked on a liberal economic policy, which is deemed to be a favorable

condition for investment and for private sector and enterprise development.

In order to implement this policy, the transitional government enacted

a number of proclamations and regulations. For instance, by Investment

Proclamation No.15/1992, the government made the rst attempt to

relax the investment climate and encourage private investment. In this

proclamation, the regional governments were also mandated to establish

their respective investment oces to promote, supervise, and regulate

investment activities (TGE, 1992). Again, by Proclamation No.41/1993,

Industry and Handicrafts Bureaus were established in regional states

to replace HASIDA (TGE, 1993). The government has continued its

commitment to create an enabling environment for investment and thus

has issued additional proclamations and amended the previous ones5.

Despite these encouraging reforms and better policy environment, the

eort has not been too fruitful in fostering MSEs, as stated in MUDC (2013).

5 See Gebreeyesus and Ambachew (forthcoming) for a detailed review of the policy and institutional evolution of MSEs in Ethiopia.

12

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

This may be the case because neither the full-edged strategy nor the

administrative organs dedicated to MSE development were in place until

the mid-1990s.

When Ethiopia adopted Agricultural Development Led-Industrialization

(ADLI) as its overall development strategy in 1995, MSE development had

started to be one of the important elements to which the strategy paid due

attention (MUDC, 2013). With the aim of improving access to nance and

counseling services to MSEs, the licensing and supervision of micronance

institutions (MFIs) was promulgated (Proclamation No.40/1996) just a year

after ADLI came into being (FDRE, 1996). This gave rise to 35 (MFIs currently

operating in dierent regions of the country, with a primary target of youth,

micro entrepreneurs, and other nancially underserved segments of the

population (NBE, 2017). Empirical evidence also arms that MFIs have

played a signicant role in the establishment and further development of

MSEs (World Bank, 2015). Having made all these arrangements, the country

introduced its rst full-edged MSEs development policy in 1997.

ii. The 1997 MSEs Development Policy

The Ethiopian MSEs development strategy formulated in 1997 – namely ‘the

Federal Micro and Small-Scale Enterprises Strategy’ (FMSES) – has served as

the basis for regional strategies (FDRE, 1997). In this strategy, the level of

paid-up capital was used to categorize businesses into the micro and small

classes, regardless of their sector type. Those small business enterprises

with a paid-up capital not exceeding Birr 20,000, excluding high-tech rms,

were categorized as micro, while those with a paid-up capital between

20,000 and 500,000 were dened as small enterprises6.

The role that could be played by MSEs was well acknowledged by the 1997

strategy. Accordingly, the sector was described as the national home of

entrepreneurship, providing the ideal environment to enable entrepreneurs

to exercise their talents. The MSE was envisaged to “generate employment

and more equitable income distribution, to activate competition, exploit

niche markets, enhance productivity and technical change, and through all

6 However, this denition has been changed in the latest strategy of the sector issued in 2011. This is discussed in the next section.

13

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

of these stimulate economic development” (FDRE, 1997, p. 9). The strategy’s

primary objective was to create an enabling environment for small and

micro enterprises. The strategy identied target groups for support and

clear criteria for prioritizing MSEs. A number of possible areas of support

to be provided by the government, including nancial assistance, training,

markets, and infrastructure, were highlighted in the strategy.

In 1998, the Federal Micro and Small Enterprises Development Agency

(FeMSEDA) was established by the Council of Ministers of Ethiopia,

Regulation No.33/1998 (FDRE, 1998), to coordinate implementation of

the strategy. This structure stretched down to regional, zone, and woreda

(district) levels. Since then, the establishment, support, coordination, and

information provisions of the MSE sector have been signicantly improved.

Sectoral policies and strategies designed after the release of the MSEs

development strategy have incorporated these basic ideas, as well as the

role that could be played by MSEs. In this respect, the Ethiopian industry

development strategy (FDRE, 2002) and the Ethiopian urban development

policy (FDRE, 2005) are notable.

Similarly, the role to be played by the MSE sector started to be explicitly

acknowledged in medium-term plans such as the Plan for Accelerated and

Sustained Development to End Poverty (PASDEP) and subsequent Growth

and Transformation Plans (GTP). Clear and ambitious targets were set for

the MSE sector in GTP I and encouraging results were registered (MoFED,

2010; NPC, 2015).

iii. The 2011 revised MSE development strategy and current developments

With a focus on achieving the ve-year targets set for GTP I and further

strengthening the development of MSEs, the government revised the MSE

strategy in 2011 to address both the framework for providing support

to MSEs and methods of implementation (FDRE, 2011). In addition to

strengthening the major points highlighted in the 1997 strategy, the 2011

revised strategy has made the following major improvements.

14

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The revised strategy has come up with a new denition of MSEs,

disaggregated by industry and service sectors. The new denition has

taken into account both paid-up capital and employment in enterprises.

This is presented in Table 2. Accordingly, when ambiguity is encountered

between manpower and total assets, total assets are used as the primary

yardstick.

Table 2: The current denition of MSE in Ethiopia introduced in 2011

Level of the enterprise Sector Human

power Total asset

Micro enterprise Industry <5 <Birr 100,000 (USD 6,000)

Service <5 < Birr 50,000 ($3,000)

Small enterprise

Industry 6-30 <Birr 1,500,000 million (USD 9,000)

Service 6-30 <Birr 500,000 ($30,000)

Source: FDRE (2011)

The strategy emphasizes that the TVET institutes play a critical role for the

development of MSEs, through acting as an extension service to industry,

so that human resources and technological development can be realized.

It widened the scope of target groups by including job creation for new

university and TVET graduates in addition to the usual emphasis on the

poor, women, and less skilled people. Furthermore, the 2011 revised

strategy has categorized the development level of enterprises as startup,

growth, and maturity. The supports to be given to MSEs are envisaged to

be specic to their level of development, at least in principle.

In order to revitalize the sector, the 2011 revised strategy emphasizes human

resource and technology development, market development, and nance

or credit service support systems. Developing the entrepreneurial ability

and technical, managerial, and marketing skills of MSE operators through

training was identied as the main area of human resource development

support.

Currently, the country is implementing the second ve-year growth and

transformation plan (GTP II), which was introduced in 2015/16 to run until

2019/20. In this plan, the MSE sector (especially manufacturing MSEs)

was given due emphasis in view of its strategic importance in minimizing

15

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

urban unemployment and achieving the vision of being a leading country

in Africa in light manufacturing industries by 2025. Clear and ambitious

targets have been set for MSEs to achieve by the end of GTP II.

In order to eectively implement this plan and incorporate the urban

food security and job creation agenda, the government has reorganized

the former federal agency (FeMSEDA) into two separate agencies. The

rst, called the “Federal Small and Medium Manufacturing Industry

Development Agency (FeSMMIDA),” was established by Council of Ministers

Regulation No.373/2016 and is accountable to the Ministry of Industry. The

other agency, the “Federal Urban Job Creation and Food Security Agency

(FUJCFoSA),” was established by the Council of Ministers Regulation

No.374/2016 and is accountable to the Ministry of Urban Development

and Housing.

The objective of FeSMMIDA is to “accelerate the expansion of small and

medium manufacturing industry in order to lay a broad base for the

development of large scale industry, maintain equitable distribution of

wealth and accelerate the transformation of agricultural-led economy to

industry led economy; and strengthen, assist and coordinate institutions

that provide support to small and medium manufacturing industry sector”

(FDRE, 2016a). As one can understand from this objective, and as stipulated

in FDRE (2016b), the development of manufacturing micro enterprises

as well as micro and small enterprises of non-manufacturing sectors is

the mandate of FUJCFoSA. It should be noted that ensuring urban food

security through productive safety and job creation programmes are

additional tasks given to the FUJCFoSA. Both agencies have their own

councils composed of high government ocials as a high governing or

policy making body.

Generally, the current MSE development policy and strategy as well as the

attention it gets from the government and other development partners

seem to be enabling. The majority of the problems encountered by

MSEs emanate from failures to fully implement the strategy and support

frameworks. However, the current changes in institutional arrangements

and the wider scope assumed by FeSMMIDA and FUJCFoSA may necessitate

some revisions to the 2011 MSE development strategy.

16

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

3. Survey methodology: sampling design, data

collection, and processing

3.1. Data and sampling

The focus of this survey is on micro and small manufacturing enterprises.

We adopt the national denition and distinguish the micro enterprises as

establishments with up to 5 employees and small rms with employment

of 6 to 30 people. Our data is a combination of a representative sample

of micro enterprises (given their sheer number) and a census of small

rms. The survey covers the 10 largest cities in Ethiopia based on their

population size7. These cities are Addis Ababa, Adama, Jimma, Dire Dawa,

Jigjiga, Mekelle, Bahir Dar, Gondar, Dessie, and Hawassa.

Due to lack of a national business directory or database, we assembled a list

of micro and small enterprises from dierent sources (federal, regional, and

city-level government agencies) to frame our population. See subsection

3.3 for further detail on our eorts to compile the list of enterprises from

dierent government agencies and the limitations of each in handling the

list.

The population frame for these enterprises in Addis Ababa was created

mainly by combining three separate administrative lists. The rst list

was compiled by the Bureau of Labor and Social Aairs through physical

visits to enterprises located on the main streets of Addis Ababa. This was

prepared for the purpose of studying employers and their labor-related

characteristics in Addis Ababa. The second list was compiled from the Addis

Ababa Trade and Industry Bureau, which has a record of all businesses that

are registered at the bureau level. The third list was obtained from the Addis

Ababa Urban Job Creation and Food Security Agency (formerly the Addis

Ababa Micro and Small Enterprise Development Agency), which conducted

a census of all businesses that received some sort of state support.

7 Since there was no accurate data on enterprise population for all cities of the country, human population was taken as a criterion to select big cities.

In doing so, we assumed that the size of human population of a city is directly proportional to its enterprise population.

17

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Having combined these lists together, further cleaning was conducted to:

1) remove rms that employ more than 50 workers from the labor data

or that have paid-up capital of more than 1.5 million birr, according to

the trade data; 2) drop duplicated rms in the three data sets; 3) remove

enterprises that are operating in non-manufacturing sectors from the list.

Table 3 reports the population and sampling distribution of manufacturing

MSEs. As can be seen in this table, our nal sampling frame for Addis Ababa

constitutes a total of 16,004 micro and small manufacturing enterprises. Of

these, 3,298 of them were small, while the remaining 12,706 were micro

enterprises.

Table 3: Population and sampling distribution of MSEs

City (Sub City) MSE Population Sample MSEs surveyed

Small Micro Small Micro

Addis Ababa (sub cities)

Addis Ketema 373 1090 373 141

Akaki Kaliti 406 1177 406 86

Arada 229 739 229 58

Bole 361 1791 361 162

Gulele 298 1144 298 100

Kirkos 191 840 191 76

Kole Keranio 229 2426 229 247

Lideta 125 520 125 59

Nifas Silk Lafto 738 1626 738 121

Yek a 348 1353 348 145

Addis Ababa Total 3,298 12,706 3,298 1,195

Sample (%) 100 9.4

Regional Cities

Adama 121 531 121 140

Bahir Dar 252 948 252 314

Dessie 89 384 89 147

Dire Dawa 126 447 126 118

Gondar 126 737 126 211

Hawassa 316 806 316 143

Jigjiga 60 120 60 44

Jimma 146 480 146 148

Mekelle 330 2449 330 850

Regional cities Total 1,566 6902 1,566 2,115

Sample (%) 100 30.6

Grand total 4,864 19,608 4,864 3,310

Sample (%) 100 16.9

18

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The frame used for the nine regional cities was based on an approach

similar to that used in Addis Ababa. The only exception was that there was

no administrative data at the regional Bureaus of Labor and Social Aairs

and thus the sampling frame was constructed using the Regional Micro

and Small Enterprise Development Agencies (ReMSEDA) and the Trade and

Industry Bureau of each region. This exercise provides us with a list of 8,468

micro and small enterprises, of which 6,902 are micro and 1,566 are small

size rms.

Combining all micro and small enterprises from Addis Ababa and the

other 9 regional cities, we obtained a total population of 24,472 registered

manufacturing MSEs. Of these, 19,608 (about 80%) are micro and the

remaining 4,864 (about 20%) are small enterprises. We then made a

decision to conduct a census of all small rms. Accordingly, all 4,864 small

size rms were surveyed; of these, 3,298 were from Addis Ababa and 1,566

were from the other nine cities in the regions. Because of the sheer number

of micro enterprises, we used a random sampling method to select the

micro enterprises to be surveyed.

As shown in Table 3, Addis Ababa accounts for 65.4% of the total population

of manufacturing MSEs in our database. We, therefore, dierentiate Addis

Ababa from the rest of the regional cities in our sampling of the micro

enterprises. About 36% of our micro sample was drawn from Addis Ababa,

while the remaining were from the regional cities. However, the sample size

within the two categories (i.e., Addis Ababa and the remaining nine cities)

was determined using the population distribution of micro enterprises

as a sampling weight. We used cities (and sub-cities in the case of Addis

Ababa) to stratify the sample and sample size. A simple random sampling

technique was then used to draw the sample of micro enterprises from

each city (and sub-city in the case of Addis Ababa). Accordingly, a total of

1,195 micro enterprises were selected from all sub-cities of Addis Ababa

and 2,115 micro enterprises were from the nine cities outside Addis Ababa.

The sampled micro enterprises in Addis Ababa represent about 10% of the

total population in our frame, while the sample from the nine other cities is

about 30% of the total population. This means the regions are purposively

oversampled when it comes to the micro enterprises.

19

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The nal cleaned data used in this study constitutes a total of 8,174 MSEs,

of which 40.5% are micro enterprises. In terms of geographical distribution,

4,493 (55%) are from Addis Ababa and the remaining 3,681 (45%) are from

regional cities. It should be noted that the total number of surveyed MSEs

is one-third of the total MSE population in the 10 largest cities and the

sample is thus quite representative.

Figure 1: Location of sample cities

In our actual dataset consisting of 8,174 MSEs, 311 enterprises were found

to engage more than 30 persons and thus they are out of the MSE category

by size8. Data on these observations are reported separately as medium-

sized enterprises throughout Chapter 4 of this report in order not to

confound the analysis based on the small-scale data. However, the medium

8 Their distribution across sample cities is presented in section 4.1 of the next chapter.

20

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

enterprises as well as data from them should not be taken as representative

or a census of all medium-scale manufacturing enterprises of the study

area, since the medium category was not included in this study by design.

Conclusions and recommendations based on this study are appropriate for

micro and small enterprises, for which the study was initially designed.

A survey instrument consisting of 15 sections and containing a rich set

of questions on dierent themes was prepared. This was programmed to

electronic versions to collect the data using Computer Assisted Personal

Interviewing (CAPI).

Table 4: Major sections in the survey instrument

Section A: General Information

Section B: Characteristics and History of the Enterprise

Section C: Characteristics of Entrepreneurs (Owners) and their Backgrounds

Section D: Growth Orientation

Section E: Employment and Earnings

Section F: Marketing and Supply Linkages

Section G: Competition, Technology, and Innovation

Section H: Investment and Finance

Section I: Business Development Services

Section J: Business Practices

Section K: Greening MSEs (Energy Use/Resource and Waste Management)

Section L: Occupational Safety and Health of Workers

Section M: Sales, Growth Opportunities, and Constraints

Section N: Cognition and Time and Risk Preference

21

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

A survey team consisting of two coordinators, seven supervisors, and 42

enumerators was formed and intensively trained on the instrument. Using

face-to-face interviews, the survey was implemented from December 2016

to May 2017 and high-quality data were collected on dierent issues of

MSEs.

3.2. Data quality control mechanisms

A number of quality control mechanisms were employed to ensure the

quality of the data. First, the programming of the survey instrument was

designed in such a way that it does not allow direct and straightforward

inconsistencies. Second, a supervisor was assigned to not more than seven

enumerators and their day-to-day operation was closely supervised by the

same.

Third, the data of surveyed rms were sent to the oce and the daily data

being collected were checked by the data manager and survey coordinators

for its consistency. At intervals, the EDRI researchers themselves checked

the quality and consistency of the data. When issues such as missing

values, inconsistent responses, or implausible numbers were discovered,

immediate feedback was given to eld sta to make corrections.

Fourth, two independent auditors were hired to verify the data quality.

For the purpose of auditing, about 600 sample MSEs were selected from

surveyed samples. A few time-invariant questions were selected from our

main survey instrument. Some new questions were also included to ensure

that the enumerator had actually contacted the respondent in person,

appropriately surveyed the rm, and asked all the questions. Using these

questions and a sub-sample of the sample MSEs, the auditors conducted a

phone survey and asked selected questions again to respondents who gave

the rst (main) interview. Then, the researchers analyzed the discrepancy

between the two responses (responses given to the enumerator and those

given to the auditor). In case signicant dierences were observed or if

22

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

there was any indication that our data collection protocol was not followed,

the name of that enumerator was recorded as a potential cheater. Then, all

rms interviewed by these potential cheaters were interviewed again by

the auditors and the responses were compared to the original completed

questionnaires. In doing so, we managed to identify interviews of about 32

rms (from the total survey) partly or fully conducted in an inappropriate

manner. Therefore, we dropped these observations from our dataset and

kept the remaining 8,174 that were proved to be interviewed appropriately.

Finally, all necessary oce-level data cleaning work was done by three

dierent experts in three rounds. In this process, when there were issues of

errors made in the eld that still needed rectication, rms were contacted

again and all necessary corrections were made.

Given all these eorts made, from the survey design to the post-survey

oce-level cleaning, we believe that we have generated high-quality and

dependable data that generates credible information.

3.3. Remarks on data handling and management practices of government oces

In the course of constructing the sampling frame and conducting this

survey, we visited dierent oces or agencies that have lists of business

enterprises compiled for some other purpose. These include oces

operating at federal, regional, sub-city, woreda, kebele (village), and one-

stop service center levels. We collected these lists and critically reviewed

their documentation processes. These sources of enterprises lists and the

nature of the data are discussed as follows because this provides a good

lesson in survey design for other researchers. Information provided in this

sub-section can also be taken as feedback to help the oces improve their

data organization and management practices.

3.3.1. The Central Statistical Authority (CSA)

CSA regularly undertakes two main surveys related to rms. These are

the “medium and large industry and electricity survey” and “small-scale

manufacturing industry survey”. Because the latter is related to our project,

23

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

we hoped that we could obtain an accurate national sampling frame

from CSA. CSA was the rst source we consulted in our eort to search

for appropriate lists of enterprises. Having made frequent contacts with

the directorate in charge of the business survey, and following our review

of small-scale manufacturing industry survey reports, we learned that

CSA does not have a complete list of MSEs using the denition of MSEs

we followed. The frame it uses for its own survey was constructed from

dierent sources, primarily from the Ethiopian Revenue and Customs

Authority (ERCA). An attempt was made to access and look at the ERCA

data for our target cities. Surprisingly, we found this source incomplete,

in that it had less than 10 small business enterprises for large cities like

Dire Dawa and Hawassa. This is denitely non-exhaustive and the source

cannot be dependable. Therefore, the research team decided to leave this

and construct its own exhaustive frame by using multiple complementary

sources, as CSA itself does for its small and cottage industry surveys.

3.3.2. The Federal Micro and Small Enterprise Development Agency (FeMSEDA)

As discussed in Chapter 2, the former Federal Micro and Small Enterprise

Development Agency (FeMSEDA) was recently split into two agencies: the

Federal Small and Medium Manufacturing Industry Development Agency

(FeSMMIDA) and the Urban Job Creation and Food Security Agency. As

the newly established agencies are too new to generate their own data,

the former FeMSEDA was contacted for the availability of MSE lists. This

source has well organized and easily accessible data updated until 2014/15

(2007 E.C) for all regions in a similar format. This source consists of basic

information about the prole of enterprises, including names, address,

TIN number, capital, labor, and size of enterprises (micro versus small). The

agency has been publishing and distributing the list annually in the form

of an MSE directory (both in hard and soft copies).

However, this source only compiles lists of enterprises established with

the support of the government, usually under the close supervision of

micro and small enterprise development oces. Those enterprises that

were established on their own without any support from the government,

24

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

other than getting a trade license and other formalities from the trade

oce, were not included in the FeMSEDA list. Such enterprises are usually

called “informal” by FeMSEDA and ReMSEDA ocers, while those formed

with their support are considered “formal.” But, one should note that this

formal-informal categorization is highly imprecise when compared to

the usual denition. Therefore, samples taken from this frame cannot be

representative. In addition, this source was updated only until 2007 E.C.

Enterprises either established or closed thereafter cannot be captured.

Most importantly, many enterprises appearing on the agency’s record

could not be found on the ground. Thus, this source cannot be used as a

single source unless it is complemented with others.

3.3.3. Regional Micro and Small Enterprise Development Agencies (ReMSEDAs)

ReMSEDA is a general name used to represent regional agencies

coordinating and supervising the development of MSEs at a regional level.

There are dierent names and structures in dierent regions. Some regions

organize this as a separate agency, while others organize it at the bureau

level. Still others merge it with other oces to organize it at the bureau

level. For instance, in the Tigray region, it is a separate regional agency,

while the Amhara region merges it with the region’s TVET oce to form

a bureau called technical vocational and enterprise development bureau.

For the purpose of the ESBD project, both the region and city/sub-city MSE

development agencies and oces were visited to collect up-to-date lists of

enterprises. In most cases, we were able to access lists of enterprises with

a complete prole and more updated data than that of the FeMSEDA. Like

the FeMSEDA’s list, all regions collect and compile basic information about

enterprises, usually on an annual basis. While many regions use a similar

format and collect similar types of information, Tigray and Amhara ReMSEDAs

further disaggregate data about the human resources involved in MSE

operations by age and education level. Regions and city administrations

were found to compile data reported from lower administrative units. In

addition, in the Amhara and Tigray regions, a regional-level census of MSEs

(including both government organized and self-initiated enterprises) was

25

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

conducted in early 2016/17 (2009 E.C) and 2014/15 (2007 E.C), respectively.

Therefore, the ReMSEDA list we collected from these regions was much

better. Not only was it up to date, but also relatively inclusive, although it

was not used as a single source for framing MSEs in our sample cities. As

far as inclusiveness is concerned, the Dire Dawa MSE development oce

tried to include enterprises established by their own motivation, in order to

provide support to such MSEs provided they were willing to register with

the bureau and obtain a MSE certicate. Otherwise, in all other regions,

support given to enterprises, as well as data that the oces compiled, is

restricted to enterprises formed with government support.

In the SNNP region in general and Hawassa city in particular, the ReMSEDA

data contains similar information as in other regions. Exceptionally, this

region generates additional information on MSEs, such as lists of model

enterprises in each sector, the lists of enterprises converted from other

sectors to manufacturing one (because of the more emphasis is given to

the manufacturing sector in nationwide planning), and, most importantly,

they a report on MSEs changed from “informal” to “formal” by bureau’s

own denition of formality. When an enterprise is changed to “formal”, the

bureau registers it as if a new rm is established.

Data collected from other ReMSEDAs were updated and compiled until

2014/15 or 2015/16 during our visit. Data from this source was easily

available in one place (region’s oce, city MSE development oce, or both)

in all cities visited, except in Jimma. In Jimma, the MSE list was collected

from the One-Stop Service (OSS) Center of each of the eight8 kebeles, as

data was not compiled at the city level to make it accessible for users.

Generally, the ReMSEDAs lists are available in Excel, contain basic proles,

and are easily accessible. However, this type of list has the following

limitations:

It is not exhaustive, as it usually excludes MSEs formed without

government support.

26

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Dead (closed) MSEs are not usually excluded from the list of the

agency. The list includes a large number of enterprises that are not

currently operating or could not go beyond registration in the rst

place.

The list contains much duplication. The duplication error is due

to the fact that the agency registers the same enterprises under

dierent names at dierent times. Usually, a rm registered in its

business name also appears by the name of the owner as a separate

rm. In the case of cooperative and partnership businesses identied

by the names of operators, any change in the name order causes

duplication.

In some cases, the title of business owners (like Mr., Mrs., etc.) is added

before their name and this makes sorting and other processing of

the list complicated. Sometimes, this causes duplication in the lists.

The micro-small categorization is not consistent across cities;

nor it is dened as per the formal denition as stipulated in the

proclamation. This is particularly true for Adama city, where all MSEs

were mistakenly labeled as micro regardless of their employment

and capital size.

MSE lists from ReMSEDA are usually written in the region’s working

language. This partly restricts external users who are not able to read

and write in that language. This is particularly true in our sample

cities located in the Tigray, Somali, and Oromia regions.

Some important information about MSEs, such as address and TIN

number, are sometimes missing.

Most of the time, the starting stage gures, especially capital, are not

updated periodically. It is very common for a rm to be operating for

a number of years before it has a level of current capital similar to its

initial level.

27

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

3.3.4. Trade and Industry Bureaus

The trade and industry bureau of each region was also identied as an

important source of enterprise lists, as all operating enterprises formally

register in this oce during the issuance and renewal of trade licenses. We

visited and collected enterprise lists from each region and city considered

in this study. This dataset commonly contains basic information such as the

name of the enterprise, owner’s name, TIN number, license number, legal

status (form of ownership), and capital. However, for the city of Jimma,

the list we collected lacks form of ownership and the capital indicator. On

the other hand, data taken from this source for Mekelle city had as many

variables as ReMSEDA, including full addresses, but with a dierent format.

As far as accessibility and data management of this source are concerned,

most cities (Hawassa, Bahir Dar, Jimma, Adama, and Dire Dawa) were found

to use a specialized software developed by the Ministry of Trade (MoT) to

handle their data. They have dedicated data managers. These cities had

updated data at the center (in one place) that could be provided for us, and

of course for any user, upon request. Some other cities simply record data

by Excel and do not have a separate data management unit (for instance,

Jigjiga). Generally, the data from this source is accessible in all cities except

Jimma.

Unlike MSEs in ReMSEDA lists, enterprises appearing on this list do not have

a problem of systematic exclusion. This is because most of the enterprises

take or renew a trade license and are recorded in the trade oce database

irrespective of how they were formed or their aliation to MSE development

oces. In order to construct an appropriate and exhaustive sample frame

for our project, the enterprise lists taken from this source were eectively

used to complement ReMSEDA data.

Though the trade and industry bureau of each region was used as an

alternative source of enterprise lists, the following limitations were

identied:

28

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

In most cases, this source does not provide a detailed address other

than kebele or sub-city, which makes it hard to locate enterprises.

Information on the type of sector only includes main sectors and

says nothing on sub-sectors in all sample cities except Mekelle and

Gondar.

For some cities, such as Gondar, the data from the center (city trade

oce) is either unavailable or outdated. Updated data are available

in sub-cities in a fragmented manner. The central data unit could

access sub-cities’ data online for some cities, such as Hawassa and

Dire Dawa. However, this was not possible for some others. In this

case, it was a must to visit all the sub-cities to collect the fragmented

information on enterprises and construct city-level data.

Disparity in data handling and management was not only observed

between cities, but also across sub-cities within the same city. In

some cases, the location of the database varies by the form of

ownership. For instance, in Bahir Dar, it was possible to get data on

enterprises with all forms of ownership except for sole proprietorship

at the center (city trade oce). Conversely, for sole proprietorship,

data can only be accessed in each sub-city.

Most data managers were unable to extract data from the Ministry

of Trade software into Excel and other user-friendly formats. After

lots of eort, some of them managed to convert it into PDF format.

Jimma’s city trade oce was exceptionally weak in this respect.

The oce used the MoT software to enter traders’ (enterprises’)

data. However, they can neither extract the recorded data from

the software, nor print it for further use. The only thing the oce

could provide us with was the entire backup of their dataset. Then,

we had to take this backup and get the technical assistance of the

MoT’s Information and Communications Technology team (the very

source of the software) to extract the data we sought from the city.

29

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Data on the capital of enterprises registered by trade oces is

highly understated. The amount of capital for the same enterprise

in the same period is dierent when it is read from ReMSEDA and

from trade bureau records.

As with ReMSEDA, language and duplication issues were also

problems with this source.

3.3.5. Federal and regional revenue and customs authorities

The revenue and customs bureau was the other alternative source that was

thought to have appropriate enterprise lists. The enterprise list collection

team sent to each city visited the respective revenue oces and collected

the data they possessed. This source was found to have very limited

variables, mainly consisting of the name of the taxpayer, TIN number, and

level of the taxpayer (labeled as A, B, and C). This source, however, has a

strong data management unit that uses uniform and specialized software

designed for this purpose. The data managers of this authority operating

at any level are relatively knowledgeable and they were able to provide

us data in any type of format, including Excel. Sub-cities or other lower

administrative units’ data could be accessed centrally for all cities, except

Adama. In the city of Adama, only the high taxpayers list was accessed at the

center, while others were obtained in sub-cities in a fragmented manner.

The main limitations of this source are:

Most of the data obtained from this source do not have a sector

indicator (e.g., Bahir Dar and Jigjiga). In some cases, even for those

lists that have a sector indicator, it is dicult to clearly categorize

their sector type. Because sector is recorded in less informative and

short manners like “timber” or “bread,” it is unknown whether the

business engages in the production of bread or in the retailing of

bread, for instance. This problem is much more common in Dessie,

Dire Dawa, and Adama.

30

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Most of the employees, particularly data managers, are unaware of

the denition of each sector, so the sector category of enterprises is

usually registered in a less than accurate manner.

Full addresses of enterprises are either not recorded appropriately

or are not easily accessible.

No size indicator (capital or labor) is included in this source, so it is

dicult to categorize businesses as large, medium, small, and micro.

Only the annual turnover (sales) value is available in this source.

As with other sources, language is another factor limiting access by

external users.

3.3.6 Lessons learned and actions required

In our review process, we learned that there is not a single national or

regional source of data that exhaustively documents the list of business

enterprises in Ethiopia. Any research that is based on any of the above

sources as the only sample frame could potentially generate inaccurate

results. Inconsistencies in data collection, organization, formats, data

updating frequencies, and management among regions, cities, and even

sub-cities within the same city are signicant. Most data sources either

lack a separate data management unit or are poorly staed by those

with inadequate technical skills, which results in poor data handling and

management practices in almost all regions. In addition, there is no link

or coordination among dierent government oces working in the same

city on the same population of enterprises in terms of data collection and

handling. Each oce collects and organizes its own information, which

results in substantial inconsistency of statistics across sources.

The absence of a complete (or accurate) national database (or business

directory) for small businesses means the state of the sector and its

contribution to the national economy, such as GDP, employment spells,

and job reallocation and destruction rates, are not accurately known. This

has allowed for some to report exaggerated gures on the enterprises

created and employment generated. Without accurate data, it very dicult

31

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

to design sound policy and also monitor the performance of the sector

over time. Thus, there is an urgent need for a national database (or business

directory) in order to improve coordination among the dierent actors (at

federal and regional levels, and also among horizontally connected public

agencies). Modern ICT systems and networking also are needed. The CSA,

in collaboration with others, needs to work hard to realize its goal of having

a national business directory. It is also advisable for the agencies listed

above to collaborate with the research community, as carefully designed,

repeated (panel) surveys can contribute greatly to responding to policy

questions.

32

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

4. Survey Results and Discussion

In this section, the main ndings of the survey are presented. The report

covers the major issues concerning MSE development. The analysis was

made using basic statistical techniques and presented in a less technical

way to make it easily understandable for policy makers, practitioners,

and other interested readers. The chapter is composed of 11 sections,

each looking at dierent dimensions of the MSE sector development.

It starts with describing the basic characteristics of sample MSEs and

entrepreneurs. Other specic topics highlighted in the objectives of this

study are presented in subsequent sub-sections. All the gures and tables

presented in this chapter are based on our survey data.

4.1. Basic characteristics of the sample enterprises

4.1.1. Size class and geographical distribution of sample MSEs

Our scientic analysis and the resulting recommendations will be

more informative if they are done in a more disaggregated way. One

way of disaggregating the data is on the basis of size class of MSEs, as

interventions needed by enterprises are usually size-specic. Therefore, we

will summarize the MSE size classes in our survey data before embarking

on the other results. The initial sample was designed to include micro

and small enterprises based on the formal denition that the country is

following. This was done using the information on size indicator on the

administrative list used for our sampling frame.

Table 5 reports the size and geographical distribution of the enterprises in

our sample. We classied the enterprises into three size categories: micro,

small, and medium enterprises. Accordingly, our database includes 3,310

(40.5%) micro enterprises, 4,553 (55.7%) small enterprises, and 311 (3.8%)

enterprises in the medium size category. The sur vey was originally designed

to include enterprises in the micro (1-5 workers) and small (6-30 workers)

size categories. However, when size category is determined on the basis of

33

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

total current employment level from our survey data, 311 (about 3.8%) of

surveyed enterprises were found to have more than 30 workers. We have

reported the medium size enterprises as a third size category. However,

as stated in the previous chapter, these 311 medium size enterprises are

not representative of medium-size rms and we cannot draw conclusions

about the medium-size sector. In contrast, the small rms were surveyed on

a census basis and the micro enterprises on a random sampling basis. Thus,

the ndings for these two categories can be considered as representative,

at least for the selected major urban centers. The survey results show that

in 2016/17, there were 4,553 small manufacturing enterprises in the 10

selected cities. The sampling exercise, on the other hand, resulted in 3,310

micro enterprises, which account for about 16.7% of the population of

micro enterprises in the selected cities.

Table 5: Geographical and size distribution of sample MSEs

City

1-6 workers

(Micro)

6-30 workers

(Small)

>30 workers

(Medium) Total

Freq. %Freq. %Freq. %Freq. %

Adama 140 4.2 113 2.5 8 2.6 261 3.2

Addis Ababa 1,195 36.1 3,035 66.7 263 84.6 4,493 55.0

Bahir Dar 314 9.5 243 5.3 9 2.9 566 6.9

Dessie 147 4.4 86 1.9 3 1.0 236 2.9

Dire Dawa 118 3.6 120 2.6 6 1.9 244 3.0

Gondar 211 6.4 124 2.7 2 0.6 337 4.1

Hawassa 143 4.3 306 6.7 10 3.2 459 5.6

Jigjiga 44 1.3 55 1.2 5 1.6 104 1.3

Jimma 148 4.5 145 3.2 1 0.3 294 3.6

Mekelle 850 25.7 326 7.2 4 1.3 1,180 14.4

Total 3,310 100 4,553 100 311 100 8,174 100

Size categories

share (%) 40.5 55.7 3.8 100

Looking at the geographical distribution of the population of MSEs, Addis

Ababa hosts twice as many as the sum total of the other nine regional

cities. The distribution of the small rms gives more insights than the

distribution of micro enterprises, as the former is census-based, while the

latter is on a sampling basis. Of the 4,553 small enterprises in operation in

34

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

2016/17 in the 10 largest cities, Addis Ababa accounted for about 66.7%

(3,035 rms). 84.6% of the medium-size rms (263 rms) are also in Addis

Ababa. This shows that the high concentration of manufacturing activities

in the capital city, Addis Ababa, is not just a phenomenon of medium and

large manufacturing enterprises (as the per the CSA reports), but is also

the case for micro and small manufacturing enterprises. Mekelle stands

next to Addis Ababa in terms of number of micro and small manufacturing

enterprises, accounting for about 14.4% of our sample (1,180 rms). Among

the regional cities, the smallest number of manufacturing enterprises was

in Jigjiga.

4.1.2. Distribution of enterprises by sub-sector

This survey tried to generate valid information on the distribution of

sample enterprises in terms of the type of manufacturing activity in which

they engaged. As depicted in Table 6, furniture and wood working, food

and beverage, and metal working workshops are the three most important

sub-sectors in which MSEs are engaged in Ethiopia. The share of these

three sub-sectors together accounted for about 63.2% of micro and small

manufacturing enterprises in urban Ethiopia (Table 6). In general, there

is similarity in sectoral distribution when disaggregating by size or city.

For example, the three sectors account for between 60% and 71% of the

enterprises in the nine cities in the regions, while in Addis Ababa, they

account for a slightly lower share (57%). On the other hand, garment,

textiles, and leather products account respectively for about 9%, 4%, and

3.6% of the MSEs in all cities.

35

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 6: Distribution of sample MSEs by sub-sector and city (%)

Economic sub-sector Addis

Ababa Adama Bahir

Dar Dessie Dire

Dawa Gondar Hawassa Jigjiga Jimma Mekelle Total (all

sample)

Furniture & wood

work 19.3 15.3 31.8 25.4 31.6 20.5 35.5 29.8 46.3 22.1 23.0

Food & beverages 18.3 40.6 21.7 29.7 18.4 23.7 16.3 17.3 21.4 20.9 20.2

Metal working

workshop 20.2 16.1 18.6 17.0 18.0 21.7 10.2 13.5 20.1 25.6 20.0

Non-metallic mineral 912.6. 18.1 2.5 0.0 17.7 5.6 9.4 7.7 1.0 15.0 11.3

Garments 8.2 1.9 12.7 14.8 5.7 21.4 8.7 20.2 6.1 10.6 9.4

Textiles 4.0 2.7 4.8 5.5 2.9 4.2 7.4 0.0 0.3 3.8 4.0

Printing 6.0 0.0 2.8 0.0 0.0 1.2 3.3 7.7 0.7 0.7 3.9

Leather & leather

products. 5.6 0.8 1.6 2.1 1.2 0.3 3.5 0.0 2.4 0.3 3.6

Chemicals 1.6 2.3 1.4 1.7 1.6 0.3 1.3 1.9 0.0 0.1 1.3

Plastics & rubber 2.0 0.0 0.2 0.0 0.4 0.0 0.0 0.0 1.4 0.1 1.2

Machinery &

equipment 0.7 0.0 0.5 2.5 1.2 0.3 2.2 0.0 0.0 0.4 0.7

Metal & woodwork 0.6 0.4 0.5 0.9 0.0 0.3 1.5 0.0 0.0 0.2 0.5

Electronics 0.4 1.2 0.5 0.0 0.0 0.3 0.2 0.0 0.0 0.3 0.3

Other 0.6 0.8 0.4 0.4 1.2 0.3 0.4 1.9 0.3 0.2 0.5

Total (%) 100 100 100 100 100 100 100 100 100 100 100

9 This also includes Bricks & other construction materials

36

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figure 2: Sub-sector distribution of sample MSEs by size of enterprises

4.1.3. Firms’ age

The survey result revealed that the MSE sector is dominated by young rms

that have operated, on average, for about ve years since establishment.

As depicted in Figure 3, about 60% of sample enterprises have been in

business for no more than ve years. Similarly, about 23.8% of the sample

MSEs have two or less years of operation. In contrast, only 16.5% of them

have been in business for more than a decade.

37

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figure 3: Age category of sample MSEs since establishment

When we look at the age distribution across sample cities, cities in Oromia

regional states (Jimma and Adama) were estimated to have about three-

fourths of their respective enterprises established in the last ve years. On

the other hand, Addis Ababa and Dire Dawa had relatively older enterprises

(see Figure 4).

The implication of having such young and less-established rms is that

both businesses are volatile, with high rates of startup and termination,

or newly established MSEs grow quickly into medium- and large-scale

enterprises and leave the MSE category. However, the FeMSEDA (2016)

reported that only 1% of MSEs managed to graduate to the medium

level over the GTP I period. Similarly, the CSA (2015) estimated the total

number of large and medium manufacturing enterprises in Ethiopia to be

only 2,758. These numbers show that the latter argument does not hold

true. A plausible reason for the boom in infant MSEs could be the volatility

of business or a peak in the entry of new rms as a reection of robust

economic opportunities or promotional activities from the government

side. However, due to the cross-sectional nature of our data, we cannot say

much about the rate of rm exit.

38

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figure 4: Age category of sample enterprises disaggregated by cities

4.1.4. Ownership type and legal status of sample MSEs

As far as the ownership of enterprises is concerned, all but three MSEs were

operated by the private sector, as expected. However, according to CSA

(2015), in large and medium size enterprises, the share of state-owned

enterprises was about 3.6%, which is relatively higher. The survey also

revealed that there are about 22 foreign-owned enterprises, most of them

in the small and medium size category. Table 7 reports the legal form of

ownership. The majority (57%) of the sample enterprises were reported

to be sole proprietorships, followed by cooperatives (14.5%). In state-

supported MSEs, however, the most common form of legal status is either

cooperative or partnership.

Table 7: Legal form of ownership

Legal status Freq. Percent

Public owned 3 0.04

Private limited 921 11.27

Share company 268 3.28

Partnership 1,130 13.82

Sole proprietorship 4,662 57.03

Cooperative 1,182 14.46

Others 8 0.1

Total 8,174 100

39

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

About 95.7% of MSEs reported that they had established the businessby

themselves from scratch. All other options, including inheritance and

purchase from others, had negligible shares (Table 8). This suggests that

entrepreneurship, particularly in the manufacturing sector, is largely a

recent phenomenon in Ethiopia, requiring a conscious strategy to promote

start-ups.

Table 8: Means of acquiring the current business

Means of acquiring Freq. Percent

Inherited 120 1.47

Started from scratch 7,820 95.67

Purchased 224 2.74

Other 10 0.12

Total 8,174 100

4.1.5. Location of sample enterprises

Specic locations of enterprises were assessed by disaggregating MSEs

located inside and outside of Addis Ababa, as this has some implications

about rms’ growth. As can be seen in Table 9, about 57% of the sample

MSEs operate outside (i.e. away from) their home in commercial and non-

commercial areas. In aggregate, about one-fth operate in industrial

zones. In this regard, there is a wide gap between Addis Ababa (28.7%) and

regional cities (9.8%), because regions, unlike the capital, are in the infancy

stage of constructing industry clusters accessible for MSEs.

40

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 9: Specic location of MSEs by cities (%)

City Home

commercial

Home non-

commercial

Outside home

commercial

Outside

home non-

commercial

Industry

zone/cluster Other Total

Adama 18.8 28.0 20.3 27.2 5.7 0.0 100

Addis

Ababa 6.7 12.5 32.5 19.3 28.7 0.3 100

Bahir Dar 1.8 20.3 28.3 30.9 18.7 0.0 100

Dessie 3.8 19.9 36.0 24.2 16.1 0.0 100

Dire Dawa 20.1 20.5 45.5 11.9 2.0 0.0 100

Gondar 2.4 22.0 44.5 22.6 8.3 0.3 100

Hawassa 14.6 12.6 38.6 22.2 12.0 0.0 100

Jigjiga 3.8 25.0 51.9 19.2 0.0 0.0 100

Jimma 25.9 15.3 37.8 18.0 3.1 0.0 100

Mekelle 5.3 13.7 55.9 16.2 9.0 0.0 100

Total 7.8 14.8 36.9 20.1 20.2 0.2 100

Firms were asked why they chose their current location. The answer given

most frequently (31.1% of the time) was that they were operating in their

current location not because they chose it; rather, it was allocated by

the government. For those MSEs who chose their location on their own,

proximity to customers (25.13%), cheaper rent (15.06%), and large area of

working premises (13.93%) were the major driving factors that dictated

their choice (Figure 5).

Figure 5: Reasons for MSEs to choose the current location

41

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

In addition to the geographical areas where rms operate, the methods of

acquiring the specic working premises (building) were also assessed. As

can be seen in Figure 6, 40% of the sample enterprises rented their current

premises from private individuals or businesses. The next largest share was

leased or rented from the government (33%). This gure is comparable with

the choice of location, particularly the state allocation option presented in

Figure 5.

Figure 6: Methods of acquiring the current business premises

4.1.6. Cooperatives: how dierent are cooperatives from others?

Cooperative businesses are run by multiple owners having diverse

interests, sometimes organized through state inducement. As a result, it

is interesting to see whether cooperative businesses are dierent from

others in their character. In our sample, there are 1,182 establishments

in the form of cooperatives. First, cooperative MSEs were asked how they

were established. As shown in Table 10, most of the cooperatives (69.7%)

were initiated by members themselves, while about 27.8% were initiated

by the government.

42

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 10: Initiator of the cooperative business enterprises

Initiated by Freq. Percent

Ggovernment 328 27.75

NGOs 26 2.2

Self/members 824 69.71

Others 4 0.34

Total 1,182 100

Being a state-induced business may not be a problem by itself as long

as the right collection of people is engaged in the right sector at the

right time. Sometimes, it is argued that state-induced cooperatives may

be formed without members’ interest and motivation or that members

forming the business might not have known each other at start-up. The

survey results revealed that about 54 (15%) of the cooperatives initiated

by the government or NGOs were organized involuntarily. Regarding

members’ familiarity with each other at the time establishment, two-thirds

of the cooperatives reported that they had known each other very well.

The remaining third had gaps in this respect.

Comparing the results across cities, Addis Ababa, Dire Dawa, and Gondar

(about 48% each) and Jigjiga (44%) have gaps in this regard, as members

do not know each other at all, barely know each other, or know each

other moderately well. Adama is better o in this regard, as about 92% of

cooperatives reported that members knew each other very well during

establishment, though this form of ownership is less common in the city

(Table 11).

Table 11: How well members know each other during establishment (%)

City Level of familiarity

Not at all Barely Moderately Very well Total

Addis Ababa 9.17 10.55 27.85 52.42 100

Adama 0 4.17 4.17 91.67 100

Jimma 0 2.5 13.75 83.75 100

Bahir Dar 6.31 5.41 9.01 79.28 100

Dessie 1.82 0 23.64 74.55 100

Gondar 3.45 24.14 20.69 51.72 100

Dire Dawa 0 20 28 52 100

Hawassa 3.45 24.14 20.69 51.72 100

Jigjiga 4.55 0 40.91 54.55 100

Mekelle 0 4.76 28.57 66.67 100

Total 5.58 7.53 20.05 66.84 100

43

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The number of cooperative members matters for the success of a cooperative

business. A large number of members may make it dicult to manage and

accommodate their diverse interests. In this study, cooperative MSEs were

reported to have on average about 7.4 and 5.6 members at start-up and

current periods, respectively (Figure 7).

Figure 7: Average size of cooperative members, by gender and period

Data on members’ dynamics over time indicated that a signicant number

of cooperative members withdrew from the business ownership. When

comparing the current cooperative members with the initial period,

their number declined by 29.5%. One reason could be members’ lack of

familiarity with each other. The less they know each other during start-up,

the less stable the number of members will be between initial and current

periods.

Pros and cons of cooperative business

Organizing MSEs in the form of cooperatives has its own advantages

and disadvantages for the businesses themselves. It is debatable which

outweighs the other. Cooperative enterprises were asked to identify

the pros and cons of being a cooperative as compared to other forms of

44

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

business. Accordingly, easy access to working premises, provided by the

government, was the primary benet of being a cooperative, which was

acknowledged by about 440 (37%) of the cooperative MSEs (Figure 8). 61%

of cooperative MSEs reported that their current location or work site was

allocated by the government, while this gure was estimated to be as low

as 29.7% for other forms of businesses. Related to this, our survey results

also revealed that cooperatives are more likely to situate in industry zones/

clusters compared to others. Therefore, the advantage for cooperatives in

this respect is evident.

Figure 8: Primary benet of being cooperative

On the other hand, 452 (38%) cooperatives reported that the diverse

interests of members was the main disadvantage of being a cooperative

(Figure 9). The diverse interests of cooperatives can lead to business

termination or withdrawal of members. As a result, there is a tendency to

discourage this form of business ownership in some cities, while others

are committed to it in forming state-induced businesses. Our survey data

revealed that cooperative businesses were more common in Hawassa,

Jimma, and Dessie, where respectively 51.6%, 27.2%, and 23.3% of the

sample MSEs were organized in this form of legal status.

45

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figure 9: Primary constraints of cooperative form of business

Other characteristics that make cooperatives dierent

The cooperative form of MSEs was compared with all other forms of

business to see whether there is any dierence in their characteristics,

using appropriate statistical tests. As presented in Table 12, generally

cooperative MSEs were found to be younger, be more necessity driven10,

have slower employment growth rates, and be less likely to hire permanent

employees, as compared to other forms of business.

Table 12: Average values of some variables for cooperatives and non-cooperative MSEs

Characteristics/variable of

comparison

Cooperative

MSEs

Non-cooperative

MSEs

Mean

dierence

P-values of the mean

dierence test

Firm age in years 4.7 6.4 1.7 0.00

Necessity driven (%) 74 64 10 0.00

Employment growth in

2015/16 10.4 17.2 6.8 0.00

Employment growth in

2016/17 3.2 5.2 2.0 0.11

Proportion of permanent

workers in 2015/16 8.3 14.8 6.5 0.00

Proportion of permanent

workers in 2016/17 8.5 15.6 7.1 0.00

10 Necessity-driven entrepreneurs are dened here as entrepreneurs who joined the business in order to balance their family and work life, due to a

lack of alternative wage-paying jobs, or post-retirement. See the following sub-section for more on the distinction between necessity- and

opportunity-driven entrepreneurs.



46

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

4.2. Characteristics of the entrepreneurs

In addition to the enterprise characteristics and the environment in which

they operate, entrepreneurs’ characteristics play a signicant role in

development of MSEs (Reeg, 2013). Entrepreneurs in the MSE context and

in this study are owners or owner-managers of the business. Our survey

results show that 94.36% of sample enterprises were managed by owners,

while only 2.87% were managed by hired managers. Therefore, the owner-

manager is more appropriate for MSEs in our case than separately referring

to these two agents in describing their prole.

Table 13 summarizes the basic demographic characteristics of entrepreneurs.

As one can see in the table, about 76% of the owner-managers of the

sample MSEs were married. The gender gap was substantial. Less than

one-fth of the MSEs were owned and managed by females. This gap

was found to be more severe in Jigjiga, where only 3.8% of the sample

manufacturing MSEs were owned and managed by female entrepreneurs.

Cultural barriers and the preference of females to engage in petty trade

rather than manufacturing are the possible reasons for this huge gap.

The average age of MSE owner-managers was 37.98 years, ranging from

18 to 89. Youth (aged 18-35 years) had a share of about 48% in owning

and managing the manufacturing MSEs. Data from CSA (2016) shows that

the share of urban youth between ages 18 and 35 is more than 60% of the

economically active urban population. This shows that the 48% business

ownership share of youth is lower as compared to the youth urban

population distribution. In terms of religion of the owner-managers, about

72.4% are Christian Orthodox, while Muslims account for about 17%.

47

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 13: Demographic characteristics of sample entrepreneurs

Variable Category Freq. Percent

Marital status

Single 1,706 20.9

Married 6,199 75.95

Divorced 113 1.38

Separated 42 0.51

Widowed 100 1.23

Total 8,160 100

Sex

Male 6,618 81.08

Female 1,544 18.92

Total 8,162 100

Religion

Christian-Orthodox 5,884 72.41

Muslim 1,384 17.03

Christian-Protestant 785 9.66

Christian-Catholic 34 0.42

Traditional 1 0.01

Atheist 9 0.11

other 29 0.36

Total 8,126 100

Age

18-35 3,895 47.74

36-65 4,110 50.37

above 65 154 1.89

Total 8,159 100

Human capital of the entrepreneurs

Educational achievements, business experience, and other skills of

entrepreneurs were assessed to measure their human capital. Table

14 presents the distribution of the summarized educational status of

entrepreneurs. As shown in this table, only 2.38% have no formal education,

while 22.8% have completed only primary school (grades 1-8). This means

about a quarter of the entrepreneurs have a maximum of primary education.

More than one-third (36%) have a high school education. More than 12%

of those in manufacturing are university graduates with at least the rst

(bachelor’s) degree level. Surprisingly, the vocational school graduates are

only about 17%, which is very low in the context of the government eorts

to massively produce TVET graduates in the hope of generating a large

pool of entrepreneurs.

48

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 14: Educational status of entrepreneurs

Educational status Freq. %

No formal education 194 2.38

Primary (1-8 grade) 1859 22.80

High school (9-12) 2951 36.20

Vocational school (10+ or level 1-4) 1383 16.96

Diploma (non-vocational) 757 9.28

First degree (BA, BSc, BEd) 883 10.83

Masters’ and PhD 116 1.42

Others 10 0.12

Total 8153 100

In addition, the total average years of education of entrepreneurs was

estimated to be 11 years. The cross tabulation of educational status of

owner-managers and rm size yielded a clear pattern on the relationship

between the two. As depicted in Figure 10, the size of the rm is positively

related to the owner’s level of education. Most of the medium size rms

are owned by more educated owner-manager while the majority of the

microenterprises are owned and managed by less educated people.

Figure 10: Education status of entrepreneurs versus rm size

49

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Our data also shows that MSE owner-managers had an average of 7.7 years

of prior business experience. The distribution of their prior occupation,

if any, is presented in Figure 11. Accordingly, about 56% of the current

business owners had been wage-employed in the private or public sector

before establishing their current business. About 28% of enterprise owners

reported that they were self-employed in a dierent business.

Figure 11: Previous occupation of entrepreneurs

In a nutshell, both the enterprise and entrepreneurial characteristics of our

sample MSEs are diverse, which has something to do with the environment

in which they operate. This has implications for incubation of entrepreneurs

and development of their enterprises.

4.3. Growth Orientation and Risk Taking Behavior of Entrepreneurs

This section discusses growth orientation and risk taking behavior

of entrepreneurs running manufacturing MSEs. The discussion on

entrepreneurial motives covers entrepreneurs’ drive, both in the context

50

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

of startup motivations and visions about the future. The risk preference of

entrepreneurs is discussed by observing the tendency to look for alternative

jobs. We also buttress this subsection by considering the engagement of

entrepreneurs in other businesses and the challenges they encounter while

running their main businesses. For ease of presentation, we disaggregate

the data by rm location and size.

4.3.1. Entrepreneurial Motivation of the MSEs

Motivation for Start ups

Entrepreneurs within the MSEs were motivated by dierent factors when

they decided to start up the business. Table 15 summarizes the major

factors that motivated the entrepreneurs to start the business, by both

rm location and rm size. The majority of entrepreneurs (more than

70%) who decided to start up businesses in the form of MSEs within the

manufacturing sector were mainly motivated to initiate the business

either due to the desire to be one’s own boss or seize the available market

opportunities. These are often called opportunity entrepreneurs in the

literature. Such types of entrepreneurs are drawn to the business not out

of desperation or being pushed out of the labor market but to exploit the

existence of protable business opportunities. The rest (about 30%) seem

to be necessity entrepreneurs, including entrepreneurs who started the

business in order to balance their family and work life (13%), due to lack

of alternative wage paying jobs (5%), for post-retirement income (3%) and

other reasons (7%).

51

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 15: Motivation for start ups (%)

Most important (top 5) reasons

for becoming an entrepreneur (%)

Location (cities) Firm size All

Addis A. Regional

cities Micro Small Medium

Wanted to be my own boss 37.8 33.7 35 36.7 35.8 36

Saw a market opportunity 35.8 33.3 31.5 36.6 40.7 34.7

Allows me to balance my family

and work life 11.4 15.6 15.2 12.3 8.1 13.3

No wage job alternatives 4.8 5.8 6.3 4.7 3.2 5.3

Post-retirement source of income 2.4 3.8 3.5 2.7 3.9 3.1

Other 7.8 7.8 8.5 7 8.3 7.7

Total no. of observations 4486 3678 3305 4549 310 8164

The business start-up motivation pattern is quite comparable between

entrepreneurs in Addis Ababa and those in the rest of cities in the regions.

However, a slightly larger proportion of entrepreneurs in the regions

(15.6%) considered business as a means to balance family and work life

than their counterparts in Addis Ababa did (11.4%). Further, the business

start-up motivation pattern was slightly dierent depending on the size

of the business. Notwithstanding the caveat that we have not generated

a representative sample of medium enterprises, we can make some

comparisons between micro, small and medium-sized enterprises. As such,

we observed that more entrepreneurs (40.7% of them) who are running

medium-sized enterprises were inuenced by the availability of market

opportunities than their counterparts from micro enterprises (31.5% of

them). Similarly, while only 8% of the entrepreneurs running medium

enterprises emphasized business as a means to balance family and work

life, more than 15% of entrepreneurs managing micro enterprises were

inuenced family-work balance. This reects the nature of micro businesses,

where the line between the household and the business is often blurred

and separating the household from businesses is dicult.

52

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Vision to Stay in Business

The entrepreneurs engaged in the manufacturing sector have diverse

visions for their businesses. We analyzed the entrepreneurs’ thoughts

on their ve-year plan for the business, the main accomplishment their

business would have to achieve to be considered a success, and the growth

potential of the business. Table 16 summarizes the ndings on the rst two

measures.

Most entrepreneurs (about 88% of them) stated that they plan to expand

their business in the next ve years, while less than 5% plan either to remain

at the same scale of operation, or downsize, or close their businesses

during the same period. The remaining 7% of the entrepreneurs do not

have concrete plans for their businesses for the next ve years. In this

regard, there is a dierence by size and location of the rms. For example,

we found that a larger proportion of the entrepreneurs in Addis Ababa

(91.1%) have a plan to expand their businesses than those in the regional

cities (83.6%) in the coming ve years. Similarly, the proportion of rms

planning to expand their business in the next ve years increases with the

size category of the rm.

53

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 16: Entrepreneurs’ vision to stay in business

Location (cities) Firm size All

Addis A. Regional Micro Small Medium

The entrepreneur’s plan for their

business for next ve years is ...

(%)



To expand the business 91.1 83.6 81.6 91.6 96.1 87.7

No concrete plans 4.5 11.4 11.7 5 1.9 7.6

To close the business 2.2 2.9 3.9 1.6 1 2.5

To remain at same scale of

operation 1.6 1.4 1.8 1.4 0.6 1.5

To downsize the business 0.6 0.8 1 0.4 0.3 0.7

Main achievement to consider

the business as successful (%) 

Continuing to grow prots year

after year 19.7 26.2 25.8 20.8 15.8 22.6

Attaining a certain level of prot 15.8 17 15.8 16.6 18.3 16.3

Expanding customer base 17.7 14.2 13.1 18.3 17 16.1

Expanding range of services &

production 20.1 7.8 9.9 17.7 18.7 14.6

Making enough to feed family 8.7 16.8 17.9 8.8 5.5 12.4

Growing to provide employment

to others 9.3 10.4 8.5 10.5 13.2 9.8

Other 0.6 0.6 0.6 0.7 0.3 0.6

Percentage of entrepreneurs

who believe in potential of the

business to grow (%)

92.9 93.4 90.4 95.3 95.8 93.4

Total no. of observations 4486 3681 3308 4548 311 8167

The entrepreneurs evaluate business success on the basis of various

accomplishments. For example, 55% of the entrepreneurs identied the

steady growth of prots, attaining a certain level of prot, or customer base

expansion as the major achievement to consider their business a success.

Entrepreneurs’ thoughts about the growth potential of their business

is another important factor in evaluating the vision they have for their

54

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

business. Overall, more than 90% of the entrepreneurs in all cities and in

enterprises of dierent sizes think that their business has good potential

for growth (see Table 16). In addition, all the entrepreneurs who think their

business has good prospects also expect that the potential growth of their

business can be realized. This shows that the majority of entrepreneurs

have optimistic prospects about the economy and their business growth.

4.3.2. Occupational choices of eentrepreneurs

In this subsection, we discuss the job preferences and occupational

choices of the entrepreneurs running the MSEs. We particularly examine

entrepreneurs’ likelihood of looking for alternative jobs, challenges they

face while running their business, and their engagement in additional

businesses.

The Likelihood of Looking for Alternative Jobs

Economic agents weigh alternatives before deciding to start and then stay

in a business. In this regard, entrepreneurs running MSEs are not dierent.

We raised various questions in relation to businesses and alternative wage

paying jobs so as to assess the level of risk the entrepreneurs are willing

to bear while running their businesses. We found out that a signicant

portion (more than 55%) of the entrepreneurs strongly agree that they

prefer owning a business to being salaried somewhere else (see Table 17).

In fact, about 86% of the entrepreneurs either strongly agree or agree that

they prefer their own business to joining wage jobs elsewhere. The pattern

does not change signicantly either with location or size of the businesses.

When we further examined the proportion of entrepreneurs currently

actively searching for wage employment, we found that only about 6% of

them are doing so. There is only a slight dierence between the regions

and size classes in this regard. As an eort to gauge entrepreneurial

propensities, we asked all respondents how they would respond to a

wage oer. A signicantly large proportion of the entrepreneurs (81%)

stated that they would not take such an oer. This is particularly the

case among entrepreneurs from medium (about 92%) and small (about

55

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

84.3%) enterprises. However, while a small proportion (about 8%) of the

entrepreneurs from medium enterprises stated that they would either

take such an opportunity or consider the nature of the job, about 25% and

16% of the entrepreneurs from micro enterprises and small enterprises

respectively said that they would do so (Table 17). We didn’t, however, nd

any sizable dierence in the proportion of entrepreneurs in Addis Ababa

and the regional cities in their tendency to take job oers elsewhere.

Table 17: Self-employment preference of entrepreneurs

Variables

Location (cities) Firm size All

Addis A. Regional

cities Micro Small Medium

The entrepreneur prefers

own business to salaried

employment elsewhere (%):

6.9

Strongly disagree 7.9 5.7 5.8 7.2 14.8 3.6

Disagree 3.5 3.6 5.2 2.4 2.9 3.4

Neither disagree nor agree 2.7 4.4 4.5 2.7 2.6 30.8

Agree 29.2 32.7 34.8 28.5 22.5 55.3

Strongly agree 56.7 53.6 49.7 59.2 57.2 6.1

The proportion of

entrepreneurs who are

currently actively searching

for wage employment (%)

4.1 8.6 8.64 4.35 4.82

Would the entrepreneur take

a job oer elsewhere? (%) 7.3

Would take it 7.3 7.3 10.8 5.1 2.6 11.7

Depends on the

nature of the job 12.8 10.4 13.8 10.5 5.1 81

Would not take it 79.9 82.4 75.1 84.3 92.3 18721

The minimum monthly salary

(in Br.) the entrepreneur

would demand to move to

wage employment

24831 9091 12065 21779 53413

Furthermore, though entrepreneurs running MSEs on average demand a

monthly salary of ETB 18,721 to move to wage employment, there is a wide

gap in such salary demands among entrepreneurs from dierent locations

56

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

and rm sizes. For instance, entrepreneurs in Addis Ababa demand an

average monthly salary of ETB 24,831, while those in the regional cities

demand an average monthly salary of ETB 9,090. Similarly, entrepreneurs

from the medium, small, and micro enterprises demand on average a

monthly salary of ETB 54,430, ETB 21,779, and ETB 12,065 respectively to

move to wage employment. This is consistent with the observation that

larger enterprises are more likely to enjoy higher prot levels in absolute

terms and hence those owners would require a larger wage premium to

move away from self-employment to wage employment.

A. Engagement in Other Businesses

We also tried to capture the level of engagement of the entrepreneurs

in terms of the number of businesses they run, the type of additional

businesses they run, and the amount of time they themselves commit to

their business. As shown in Table 18, the majority of these entrepreneurs

(about 83%) do not have any additional businesses that they are currently

running. Only about 15% of the entrepreneurs are currently running one

additional business, while a very small proportion (2%) are running two

additional businesses. In this regard, we didn’t observe any signicant

dierence on the basis of location.

However, we found that a relatively larger proportion of entrepreneurs

from medium enterprises (about 26%) are currently running one or two

additional businesses as compared to 14% and 18% of entrepreneurs from

micro and small enterprises respectively. Among entrepreneurs who have

an additional business, we found that the majority (about 77%) are running

businesses dierent from the line of business of the main one, while a

smaller proportion of the entrepreneurs (about 21%) have additional

businesses similar to the main one.

57

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 18: Engagement in other businesses

Variables

Location (cities) Firm size

All

Addis A. Regional Micro Small Medium

The number of additional

businesses the entrepreneur is

currently operating (%)



Zero 82.3 83.5 86.1 81.2 71.4 82.8

One 15 14.1 12.5 15.8 20.3 14.6

Two 2.1 2 1.3 2.3 6.1 2

More than two 0.6 0.4 0.1 0.7 2.2 0.6

If the entrepreneur has additional

business, what type is it? (%) 

Similar to the main

business 23.7 17.2 15 22.4 30.6 20.5

Dierent from the main

business 72.4 81 81.9 74.8 67.1 76.7

Both similar and dierent 3.9 1.8 3.1 2.8 2.4 2.9

Number of observations with

additional business 616 606 419 718 85 1222

Average no. of hrs. the

entrepreneur personally works in

the business in a typical week

52.4 52.5 51.5 53.2 52.3 52.4

Total no. of observations 4492 3681 3310 4552 311 8173

The observation that nearly 30% and 20% of medium and small enterprises

respectively are running more than one business is reective of the nature

of business organization in many developing countries. As opposed to

expanding their businesses vertically by employing more workers, many

entrepreneurs choose to venture into another business. This is partly a

risk mitigation eort through diversication of businesses ventures and

partly a strategy to operate outside the purview of the legal system to

escape regulation and taxation. Horizontal or lateral expansion, however,

has serious consequences for the growth and transformation potential of

micro and small businesses. Productivity and growth-enhancing gains for

economies of scale in transactions involving input, production and output,

as well as marketing and distribution, are lost when entrepreneurs diversify

into dierent businesses. Later expansion also increases costs associated

with the monitoring and management of workers and inventories.

58

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Further, we found that, on average, an entrepreneur spends 52.4 hours of

his/her working time on business in a typical working week. The gure is

comparable whether the entrepreneur is in Addis Ababa or the regional

cities. However, we observed that, on average, while entrepreneurs from

the micro enterprises work an hour less than the average, those from

the small enterprises work an hour more than the average. Overall, the

entrepreneurs within the MSEs are working much more than the common

40 hours a week requirement in the public sector.

4.4. Job creation and employment dynamics of MSEs

Ever since Birch’s (1979) classic work on the contribution of small rms

to employment generation, there is a growing body of literature on the

economic role of small rms. Most of these studies show that small rms

employ a large share of workers in developing countries. Consequently,

MSE development has become a key priority in many developing countries,

including Ethiopia. In this sub-section, we presented some ndings

showing the employment dynamics of MSEs in Ethiopia.

4.4.1 . Employment dynamics by type of employees

Table 19 reports the employment generated by the MSEs by type of work and

year. The total sum of employment generated by the MSEs in our sample in

2016/17 is 72,584, of which 72% are paid workers. Between the initial year

and 2016/17, total employment has grown from 50,633 to 72584, which

amounts to 43.3% growth. However, the initial year is not very informative,

given the dierence in the starting year of the MSEs. When looking at the

last two years (that is, between 2014/15 and 2016/17), total employment

grew by about 14.7%, while the last year’s growth (2015/16 to 2016/17)

was stagnant and in fact exhibited some decline. At the initial year of

business, the total number of working owners accounts for near to half of

the total jobs created by the MSEs. However, the share of working owners

declined over time and reached about a quarter in 2016/17. In contrast, the

share of paid workers increased from about half to nearly three-quarters of

59

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

employment. This suggests that the MSEs are increasingly generating paid

employment beyond themselves and their family members.

Table 19: Employment dynamics by type of workers over time (for our sample MSEs)

Initial period 2014/15 2015/16 2016/17

Worker type No. Share (%) No. Share (%) No. Share

(%) No. Share (%)

Working owners 24212 47.8 18208 28.8 19120 26.2 18633 25.7

Paid workers 25001 49.4 43189 68.3 51895 71.1 52244 72.0

Unpaid workers 1215 2.4 1106 1.7 1361 1.9 1445 2.0

Apprentices 205 0.4 752 1.2 562 0.8 262 0.4

Total engaged

people 50633 100 63255 100 72938 100 72584 100

4.4.2. Employment dynamics by type of employees and size category of MSEs

Beyond the total number of jobs created by MSEs, it is also relevant to assess

whether the job creation potential of MSEs is dierent based on their size.

For instance, out of the 72,584 jobs created by our sample rms in 2016/17,

about two-thirds were created by small enterprises. However, this might

be misleading given that the microenterprises in our database are based

on a sample (only 17% of total population) and the small size enterprises

on a census basis. To estimate the total employment generated by the

microenterprises, we multiplied the average employment of this size class

(which is 3.3 for year 2016/17) by the total number of microenterprises in

our population frame (which is 19,608). This gives us about 64,706 jobs in

the microenterprises size group (see Table 20). As mentioned earlier, the

medium size sector is also not representative of its size class. We are not,

thus, attempting here to estimate the population. In sum, we can conclude

that the micro and small size enterprises (employing 1-30 workers) in the

10 largest cities in Ethiopia are estimated to have generated about 113,705

jobs. This estimation leads to a rise in the share of the microenterprises

employment generation to about 57% of the total employment generated

by the MSEs.

60

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figure 12: Total amount of employment created by population of enterprises in sample cities by size

and time

*estimated employment for the population of microenterprises

Table 20 gives employment generated by size category and location for

the year 2016/17. As indicated above, we have adjusted the employment

generated in the microenterprises by their weight in the sample. The table

shows that the rms in Addis Ababa account for about two-thirds of total

employment created in the manufacturing MSEs sector. The nine cities in

the regions account for one-third of employment, which is consistent with

the distribution in the number of enterprises in the two locations. Addis

Ababa’s share in employment increases with the size category of rms.

Table 20: Total employment distribution by rm size and location in 2016/17

Total

employment

Micro Small Medium Total

No. Share

(%) No. Share

(%)

Addis Ababa 42966 65.4 33984 69.4 10738 84.4 87688 68.8

Regional

cities 22778 34.6 15015 30.6 1979 15.6 39772 31.2

Total 65744 100 48999 100 12717 100 127460 100

61

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

4.4.3. Nature of Employment in MSEs

The other aspect of MSE employment is whether the employment they

create is permanent or temporary. As shown in Table 21, the proportion

of permanent employment is small relative to temporary employment,

although there is an increase from 17 percent in the initial year to 22.7

percent in 2016/17.

A great chunk of the employment created is temporary in the sense

that the majority of their employees do not have a written employment

contract with a contract period longer than six months. The temporary

work accounts for 77%-83% depending on the year under consideration.

Table 21: Employment dynamics by nature of employment over time

Nature of

employment

Initial year 2014/15 2015/16 2016/17

Number % Number % Number % Number %

Permanent

workers 8579 17.0 13707 21.7 15639 21.4 16481 22.7

Temporary

workers 41856 83.0 49504 78.3 57305 78.6 56181 77.3

Total 50435 100 63211 100 72944 100 72662 100

4.4.4. Employment dynamics by gender of workers

Table 22 looks at the employment dynamics by gender. MSEs employ more

male workers than female workers – 85 percent male in the initial year.

There is a slight improvement in the most recent years as the proportion of

female workers increased from 15 percent in the initial year to 21 percent

in 2016/17.

Table 22: Employment dynamics by gender over time

Type of workers by

gender

Initial year 2014/15 2015/16 2016/17

Number % Number % Number % Number %

Female workers 7492 15 11678 19.5 12555 17.2 14850 20.5

Male workers 42943 85 51533 81.5 60389 82.8 57812 79.5

Total employment 50435 100 63211 100 72944 100 72662 100

62

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

4.4.5. Educational level of MSEs’ Workers

Regarding human capital of workers, two observations are relevant (see

Table 23). First, overall human capital accumulation has shown slight

improvement over time. Second, most of the workers (more than 87

percent) in the MSE sector are high school graduates and below as of

2016/17. Thirdly, TVET and college graduates constitute a small proportion,

respectively 7 and 6 percent of the workforce.

Table 23: Educational Background of MSE workers over time

Status

Initial year 2014/15 2015/16 2016/17

No. %No. %No. %No. %

Below high school 32596 64.6 39118 62.0 44626 61 44085 60.67

High school

graduates 11698 23.2 16074 25.4 19079 26 19216 26.44

TVET completed 2999 6.0 4306 6.8 4869 7 4993 6.88

Diploma and above 3142 6.2 3713 5.8 4370 6.00 4368 6.01

Total 50435 100 63211 100 72944 100 72662 100

4.4.6. Labor Turnover

One manifestation of good working conditions is a lower level of labor

turnover. To shed light on this, enterprises were asked how many employees

left the enterprise of their own will and how many were red by the rm,

out of the total employees they hired 12 months prior to the survey period.

The summary results are reported in Table 24.

Table 24: Average employees’ turnover by location

Addis

Ababa Regional cities Total(All

sample)

How many employees did the establishment

hire in the past 12 months? 6.3 3 4.82

How many employees left this business in

the past 12 months of their own will? 4 2 3.07

How many employees were red from this

business in the past 12 months? 0.6 0.33 0.48

63

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

The results reported in Table 24 indicate that, on average, rms located in

Addis Ababa hired six employees during the 12 months prior to the survey

month, while rms located in regional cities hired only three employees.

During the same period, on average four employees from rms operating

in Addis Ababa and two from rms operating in regional cities quit the job

of their own will, reecting a high rate of labor turnover in MSEs operating

both in and outside Addis Ababa. According to the respondents (rm

managers), for more than four-fth of rms operating both in Addis Ababa

and other regional cities, the most common reason for workers to quit their

job of their own will is to look for a better salary (see Table 25). This result is

similar to the rm level study conducted by Kumar (2011). Another strand of

literature, such as Nagaya (2015), indicated lack of retention strategies, low

level of employee motivation, lack of career development opportunities

and poor work environment as the root causes for labor turnover.

Table 25: Most Common Reasons for Workers to Quit (%)

Reasons for labor

turnover

Addis

Ababa

Regional

cities Total sample

For better salary 80.74 81.86 81.23

For decent work 9.90 10.10 9.99

Residence change 4.47 3.42 4.01

Starting own business 2.97 3.03 3.00

Migration 0.74 0.94 0.83

Death 0.05 0.03 0.04

Sickness 0.44 0.09 0.29

Other 0.7 0.53 0.62

Total 100 100 100

No. of observations (N) 4493 3681 8174

Firms operating in Addis Ababa and regional cities red 0.6 and 0.3

employees on average respectively during the same period. According

to the rm managers/representatives, the rst and second most common

reasons for workers to be red are low productivity and poor discipline (see

Table 26).

64

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Table 26: The Most Common Reasons for Workers to be Fired

Most common reasons

for a worker to be red

Addis Ababa Regional cities Total

No. Percent No. Percent No. Percent

Low productivity 1,931 48.5 1,492 44.76 3,423 46.8

Poor discipline 1,578 39.6 1,538 46.14 3,116 42.6

Business contraction 459 11.6 301 9.04 760 10.4

other 11 0.3 2 0.06 13 0.2

Total 3979 100 3333 100 7312 100

4.5. Earnings, Working Conditions and Occupational Safety of Workers

The World Bank reports that an additional 600 million jobs need to be

created by 2020 to keep up with demographic changes and population

growth in low- and middle-income countries, just to keep the ratio of

employment to working-age population constant (World Bank, 2016).

However, there is also a growing consensus that, in order to reduce poverty,

we do not just need more jobs but also, and more importantly, better paid

jobs and decent working conditions11.

Thus, to be able to contribute to poverty reduction, it is not only important

whether MSEs generate new jobs, but also whether these jobs oer decent

working conditions and fair compensation for their employees. More

importantly, a productive job has to oer sucient earnings to permit the

workers and their dependents a level of consumption above the poverty

line. The Ethiopia government in its successive growth and transformation

plans has emphasized the importance of the establishment and growth

of manufacturing MSEs that create productive jobs for non-high school

graduates and other low-skilled segments of the society, to reduce poverty

and make its structural change more inclusive. This section briey presents

the ndings of the survey on the earnings, working conditions and

occupational safety of workers hired in these manufacturing MSEs.

11 According to estimates from the International Labour Organization (ILO), nearly half of all workers worldwide still live below the

USD 2 a day poverty line (ILO-IMF, 2010).

65

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

4.5.1. Earning of workers

Table 27 reports the average monthly earning of production workers in the

MSE sector. It indicates that, on average, MSE production workers received

a basic after-tax salary of ETB 2069 ETB (USD96)12 in 2016/17, which is much

higher than the 2 USD per day (60 USD per month) poverty line and the 49

USD real monthly per capita income of the country13. As can be inferred

from Table 27, the average monthly salary for the MSE workers also shows

a signicant improvement over time and was higher than both the poverty

line and the national real per capita income. However, since we do not

have data on the number of dependent family members of each worker,

we cannot comment on whether the MSEs oer sucient earnings to their

worker to permit the workers themselves and their dependents a level of

consumption above the poverty line.

Table 27: Average Monthly Earning of Production Workers (in ETB)

Year Mean Standard deviation Number of observations (N)

2014/15 1669 1197 6,436

2015/16 1899 1611 7,155

2016/17 2069 1698 7,162

Gender-disaggregated analysis of the earnings data uncovers a statistically

signicant dierence between male and female production workers’

average monthly earning. As shown in Table 28, the average monthly

earning of male production workers is 20% to 25% higher than their female

counterparts, depending on the year. Strikingly, the wage gap between

the male and female production workers has widened over time. The

highest gender gap is registered in Mekelle, while the lowest gender gap is

registered in Jigjiga and Dire Dawa in 2016/17.

More strikingly, the average monthly earning of the women workers in

Bahir Dar, Dessie, Gondar, Jimma and Mekelle is much lower than the 2

USD per day poverty line. Hence, the jobs created at the manufacturing

MSEs do not help female production workers escape from poverty in any

of the study cities. This calls for the attention of policy makers and other

stakeholders who promotes gender equality14. However, this does not

12. A UN average exchange rate of 1USD=21.5 USD is applied

13. According to world development indicator, the per capita income of Ethiopia in 2016 was 511 USD (World Bank, 2017a)

14. This results do not control for the skill (education and experience) gap of the male and female workers.

66

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

mean that manufacturing jobs aggravate the gender gap. Rather, they

contribute substantially in reducing gender inequality According to world

development indicator, the per capita income of Ethiopia in in terms

of earning (see also Getahun and Villanger, 2017). The average monthly

earnings in all the cities under study is much higher than the country’s

minimum wage of ETB 570.

Table 28: Average Monthly Earning of Production Workers Disaggregated by Gender and cities

Year 2014/15 2015/16 2016/17

City Male Female Di Male’s Female Di Male Female Di

Adama 1620 1297 322** 1879 1410 469*** 2095 1512 583***

Addis

Ababa 1906 1548 358*** 2129 1714 415*** 2309 1859 449***

Bahir Dar 1484 1087 396*** 1617 1171 445*** 1819 1270 548***

Dessie 1344 828 515*** 1475 882 593*** 1588 927 660***

Draie

Dawa 1492 1151 341*** 1692 1322 369*** 1789 1391 398***

Gondar 1429 961 468*** 1587 1074 512*** 1747 1140 607***

Hawassa 1792 1654 138 2071 1729 341** 2217 1850 367***

Jigjiga 1825 1529 295 1960 1688 271 2066 1806 259

Jimma 1404 832 572*** 1513 866 647*** 1592 890 702***

Mekelle 1528. 853 675*** 1751 899 851*** 2003 963 1,309***

Total 1746 1455 291*** 1957 1596 388*** 2138 1727 411***

N5964 2882 6600 3266 6597 3281

Note: **** indicates the value is statistically signicant at 1% and ** indicates the value is statistically signicant at 5%

The average monthly earning also signicantly varies across the study

areas. As shown in Table 28, in 2016/17 the rst and second highest

average monthly earnings were registered in Addis Ababa and Hawassa,

respectively, while the rst and second lowest average monthly earnings

were registered in Dessie and Jimma, respectively. This clearly illuminates

substantial city-level heterogeneities in terms of production workers’

earning. However, we note that this may partly described by the variation

in living cost across the cities.

67

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Disaggregation of the average earning of workers by the size of the

enterprises also revealed important rm heterogeneities. As depicted in

Figure 14, on average, medium-sizes rms pay higher compensation to

both female and male workers than do small and micro rms. The positive

correlation between the size of the enterprise and the compensation might

illuminate a trickle-down eect of rm growth or a variation in the workers’

skill level.

Figure 13: Average monthly earnings of Workers in Birr Disaggregated by Size of the Enterprise

4.5.2. Occupational Safety and Health (OSH) of workers

Ethiopia has ratied 22 ILO conventions and 12 technical conventions,

including the Occupational Safety and Health Convention (No.155). The

Minister of Labor and Social Aairs (MOLSA) issued an OSH directive in

2008 pursuant to Article 98(3), 102(1), 170(1) of the Labor Proclamation.

The OSH directive requires workplaces employing more than ten workers

to establish occupational safety and health committees (OSH committees).

The directive was intended to make workers aware of occupational

safety and health issues in the workplace (OSH, Section 60(1)). The OSH

68

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

directive sets out the details of the employer’s responsibilities, including

the provision of personal protective devices (OSH, Part 2), procedures to

prevent accidents, hazard and risk reduction measures (OSH, Part 4, 5),

and occupational exposure limits to chemicals, heat, noise, and vibrations

(OSH, Schedule 3, 4 , 6, 12).

Systematic study of workers’ occupational health and safety (OSH) has

been limited in Ethiopia, with the exception of administrative reports.

This subsection presents survey ndings on OSH, which are captured in

Table 29. The survey shows an average of 1.6 work-related accidents in the

surveyed enterprises in the six months prior to the survey month, with

a higher occurrence of work-related accidents registered in enterprises

located outside Addis Ababa (1.7) relative to enterprises located in Addis

Ababa (1.5). This is harmful for both workers and enterprises. A report by

WHO revealed that work-related accidents inhibit enterprises’ growth and

productivity by increasing recruitment and training costs for replacement

workers, reduce productivity of injured workers, and compromise product

quality (Burton and WHO, 2010).

The survey response shows that 70 per cent of sample enterprises have

made their workers aware of the general occupational safety and health

issues in the factory. In this regard, there is signicant variation between

rms located in and outside Addis Ababa; respectively 74% and 65% of

rms have made their workers aware of the general occupational safety

and health issues in the factory. One has to use caution in interpreting

these numbers, as they are self-reported and might be an overestimation.

About 65% of sample enterprises reported having provided workers with all

necessary personal protective clothing and equipment. The proportion of

enterprises located in Addis Ababa that provide workers with all necessary

personal protective clothing and equipment is again signicantly higher

than those enterprises located outside Addis Ababa. The majority of rms

located both in and outside Addis Ababa also stated that noise level,

temperature and ventilation at their workplace is acceptable and that their

work place is adequately lit, clean and tidy.

69

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

However, the proportion of enterprises with acceptable levels of noise,

temperature and ventilation, and a clean and tidy workplace, is signicantly

higher in the capital city than in regional cities. Among the MSEs that use

chemicals and other hazardous substances, only 16 % of rms eectively

trained workers who work with chemicals and hazardous substances. In

addition, less than a quarter of enterprises provide adequate washing

facilities and cleansing materials in the event of exposure to hazardous

chemicals. About 70% of MSEs have made workers aware of the general

occupational safety and health issues in the factory –again with signicant

variation between rms located in and outside Addis Ababa.

Table 29: Occupational Safety and Health indicators disaggregated by the location of rms

Location of enterprises Addis

Ababa

Regional

Cities Total P- value*

Firms whose workers are familiar with

proper waste handling & emergency

procedures

72.1% 63.3% 68.1% 0.000

Enterprises that made workers aware of

general occupational safety and health 74.3% 65.1% 70.1% 0.000

Eectively trained workers who work with

chemicals & hazardous substances 19.8% 12.6% 16.3% 0.000

Enterprises that provide adequate

washing facilities and cleansing material 33.2% 14.1% 24.2% 0.000

Enterprises that provide workers with all

necessary personal protection clothes 72.1% 56.8% 65.2% 0.000

Enterprises with acceptable temperature

and ventilation 53.6% 50.5% 52.2% 0.006

Enterprises with an acceptable level of noise

64.4% 57.3% 61.2% 0.000

Enterprises with adequately lit workplace 78.4% 78.4% 78.4% 0.964

Enterprises with clean and tidy workplace 69.8% 78.1% 73.57% 0.000

Average work-related accidents in the last

six months 1.7 1.5 1.6 0.000

Number of observations (N) 4493 3681 8174

Note : $ the P-value associated with a t- test with unequal variance where location is the grouping variable

Disaggregation of the occupation and safety status of workers by size of

the rm also unveils important rm heterogeneities. Measured in terms of

all 10 occupation and safety standards, medium-sized rms perform better

than micro and small rms, and small rms perform better than micro size

70

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

rms (see Table 30). For example, more than four-fths of medium-sized

enterprises, but only about three-fths of micro enterprises, make their

workers aware of the general occupational safety and health standards

and provide their workers with all necessary personal protection clothes.

Likewise, more than half of the medium-sized enterprise but less than a fth

of the micro enterprises provide adequate washing facilities and cleansing

material. Compared to micro and small enterprises, a relatively larger

proportion of medium-sized enterprises label and/or store chemicals and

hazardous substances properly. In addition, the workplaces of the medium-

sized enterprises are tidier and cleaner and more likely to have adequate

ventilation and temperature than their micro and small counterparts.

However, there seems to be no statistically signicant association between

the size of the enterprise and the number of work-related accidents the

enterprise experienced.

Table 30: Occupation and safety status of workers by size of enterprises

Indicator Micro Small Medium Total

Firms whose workers are familiar with proper

waste handling & emergency procedures 62% 72% 77% 68%

Average number of work-related accidents in the

last six months 0.15 0.17 0.2 0.165

Enterprises which made workers aware of

general occupational safety and health 63% 75% 82% 70%

Eectively trained workers who work with

chemicals & hazardous substances 13% 17% 38% 16%

Main Features of Micro and Small Manufacturing Enterprises in Ethiopia

Figures