Looking beyond the visible : how food value chain developments affect its neglected stakeholders : case study of the female Injera bakers in the Teff value chain in Addis Ababa, Ethiopia

(1)

LOOKING BEYOND THE VISIBLE: HOW

FOOD VALUE CHAIN DEVELOPMENTS

AFFECT ITS NEGLECTED STAKEHOLDERS

Case Study of the Female Injera Bakers in the Teff Value Chain in

Addis Ababa, Ethiopia

Diede Smith

International Development Studies

(2)

(3)

Looking Beyond the Visible: How Food Value Chain

Developments Affect Its Neglected Stakeholders

Case Study of the Female Injera Bakers in the

Teff Value Chain in Addis Ababa, Ethiopia

MSc International Development Studies 2018-19

Graduate School of Social Sciences

Diede Smith

10753044

July 8

th

_{, 2019}

Word-count: 24995

Supervisor: Drs. Josh K. Maiyo

(4)

ABSTRACT

Women and their labour are often not adequately reflected in policies that affect them. This also happens in food value chain (FVC) developments. Women play an important role in realizing the developments in FVCs, but their labour is often neglected in related policies. The female injera bakers in Addis Ababa, Ethiopia, have suffered a similar fate: they have been neglected in the policies on improving the teff value chain. The purpose of this research was to look beyond the visible and to investigate how the developments in the teff value chain over the last five years have affected the visibility and possibilities for empowerment of the female

injera bakers in Addis Ababa. The question that led this research was: ‘How have the changes

in the food value chain of teff affected the visibility and empowerment of female injera bakers in Addis Ababa, Ethiopia?’. This research relied on mixed methods, including twenty-three semi-structured interviews, forty surveys and participant observation.

It was found that the developments in the teff value chain were overall more positive in the formal than in the informal sector. The female injera bakers in the formal sector saw increased opportunities for empowerment due to the changes. Since the relation between the chain and the sectors are mutually constitutive, meaning that they are in constant interaction and constantly influence each other, this thus simultaneously led to positive developments in the

teff value chain. The female injera bakers in the informal sector experienced increased

invisibility of their labour, and less empowerment opportunities due to the changes in the chain. These negative effects run along the lines of the formality of the labour, as well as gender, class, age and ability. By neglecting the female stakeholders in the developments of the teff value chain, existing inequalities increased and the most vulnerable have been further marginalised. The policy recommendations coming forth from the analysis of this research is fourfold. First, it is argued that informal bakers must get the opportunity to register their businesses at the local

woreda so their businesses can be acknowledged by the public. Also, they should be able to get

quality marks which can prove the legitimacy of their labour and the products they make. All without taxes being held from their profits which barely make a living wage. Second, the existing labour rights should be promoted by either the Ethiopian government or aid organisations. This will lead to greater knowledge of their rights as labourers, as well as an increased sense of collectivity which can lead to a better bargaining position vis-à-vis their

(5)

clients. Third, women’s representation in leadership position in the government must be increased in order to open up the public discourse on gender equality, to show that women can be equal leaders and labourers to men and to eventually trickle-down into society and thereby to reduce gender inequality in Ethiopia. Fourth, the Ethiopian government should protect the

injera bakers by setting a minimum/controlling the price of injera, subsidising injera for the

local consumers and support the bakers with human and material resources.

Key-words: Food value chain (FVC), invisible work, gender, empowerment, teff value chain, Addis

(6)

ACKNOWLEDGEMENTS

First, I would like to thank my Addis Alem Tekalign for being an incredible research assistant and friend, without whom I could not have done this research. Next, I would like to thank my friend Henok Kassahun who has shown me the ways of Addis Ababa and to whom I could always turn for support. Also, I would like to thank all my friends for continually supporting me when I was in the field and during the writing process of this thesis. Furthermore, I would like to thank Josh Maiyo for being my supervisor. Next, I would like to thank Mirjam Ros-Tonen for agreeing to serve as a second reader. Finally, I would like to thank all respondents who have so eagerly welcomed me inside their businesses, their homes and more importantly into their personal lives.

(7)

LIST OF FIGURES AND TABLES

Figure 1. Conceptual Scheme 22

Figure 2. Map of Ethiopia 36

Figure 3. Map of Addis Ababa by sub-city 36

Figure 4. A street near Mercato with a man carrying buckets, shops 36 Figure 5. A passage with souks in one of the old ‘villages’ of Kirkos 37

Figure 6. One of the main roads in Yeka 38

Figure 7. A street running through one of the villages in Lideta with small shops 38

Figure 8. Injera 45

Figure 9. Three-stone fire supporting a mitad (clay pan) 45

Figure 10. Electric injera mitad 45

Figure 11. Injera being transported on top of a man’s head 50

Figure 12. Education of the injera bakers in the formal and informal sector 58 Figure 13. Household income of the injera bakers in the formal (left) and 65 Figure 14. Age of the injera bakers in the formal sector and informal sector 65

Table 1. Operationalisation 23

Table 2. Overview of data sources 26

Table 3. Overview of codes and key themes 29

(10)

LIST OF ACRONYMS

ADLI Agricultural Development Led Industrialisation

BCE Before the Common Era

ERPDF Ethiopian Peoples’ Revolutionary Democratic Front

FDI Foreign Direct Investment

FGM Female Genital Mutilation

FVC Food Value Chain

GDP Gross Domestic Product

GEM Gender Empowerment Measures

GNP Gross National Product

LDC Least Developed Countries

LIC Low-Income Countries

MoWA Ministry of Women Affaires

MoWCYA Ministry of Women, Children and Youth Affaires MPI Global Multidimensional Poverty Index

SME Small and Medium Enterprise

SSA Sub-Saharan Africa

(11)

1. INTRODUCTION

The neglect of women’s labour appears throughout markets and countries. Much of the labour women carry out is rendered marginal, inferior and invisible due to the conceptualization of the economy in binary and gendered terms. As a result, women’s labour is not adequately reflected in projects, policy recommendations and developments that affect these women. A similar situation is evident in food value chain (FVC) developments. While it is widely known that female stakeholders in FVCs are crucial for its development, their labour is often not taken into account in policies that aim to achieve these developments. As a result, many of the intended FVC development outcomes cannot be achieved, and the overlooked women are obstructed in their own development by the unintended consequences of their overlooked status. This thesis looks beyond the visible in food value chains in order to investigate how women’s labour is overlooked in policies, and how the characteristics, dynamics and processes within FVCs affect the socio-economic position of these women. This research focuses on the case of female injera bakers in Addis Ababa, Ethiopia, and how the changes in the teff value chain affect their visibility and possibilities for empowerment. Thereby, it looks at the structuring elements in the value chain and how it affects the female injera bakers, and vice versa.

This chapter first defines the research problem of women’s neglected labour globally and in Ethiopia (see 1.1.1), the role of FVCs in societal developments and the integration of women within these developments (see 1.1.2) and the teff value chain and the role of women (see 1.1.3). It then explains the purpose of this research (see 1.2) and the main research question and sub questions that have led the research process (see 1.2.1). This chapter ends with an overview of the thesis structure per chapter (see 1.3).

1.1 Problem Statement

1.1.1 Women’s Neglected Labour

Conventionally, the economy is understood as “the manner in which resources and human labour are combined to produce and distribute material goods and services” (Bryson 1996: 208). With the neoliberal turn the capitalist market began to be constructed in ways that relegated women’s labour to the family- and household economy and assigned men’s labour to the market economy (ibid.: 206). This conceptualization of the economy in binary and gendered terms has

(12)

resulted in rendering the (unpaid) labour executed by women, mostly in agriculture, food production and household activities, as marginal, inferior and invisible (ibid.: 208-210). This invisibility constitutes itself in countries’ Gross National Product (GNP) and other statistics, and hence (in)directly in projects and policies that affect these women (ibid.: 210).

In Ethiopia’s labour market, the gender gap includes both a horizontal and vertical segregation: women are dominant in different sectors than men, and women often obtain lower valued jobs than men (Baughn 2006: 390; Alibhai 2015). For instance, men dominate sectors such as engineering, tourism, transport-services and business services, while women dominate sectors such as beauty, food-services and communication (Alibhai 2015: 2). Next to this, the male-dominated sectors have higher returns than the female-male-dominated sectors (ibid.: 2). Also, women perform more than double the amount of unpaid work as men (340 minutes versus 141 minutes per day) (CIMMYT 2018: 2).

The Ethiopian government has, in cooperation with UN Women (2014: 39), articulated that “making women’s lives visible and integrating women’s economic power and activities into formal economic planning” is key in its aim to foster gender equality. General institutional mechanisms in Ethiopia concerning women’s labour are relatively advancing, as with the establishment of the Ministry of Women’s Affairs (MoWA), the Ministry of Women, Children and Youth Affaires (MoWCYA), and the inclusion of Gender Responsive Budgeting by the Gender Directorate of the Ministry of Finance and Economic Development (ibid.: 13). However, women’s (in)formal monetary and non-monetary labour, activities and roles that contribute to the society are still not commonly addressed in Ethiopian governmental policies and statistics (ibid.: 38).

Most development projects in Ethiopia concerning agriculture, women and FVCs focus on rural areas/upstream segments of FVCs and do not have an explicit gender (mainstreaming) strategy. The UNDP (2018b: 5) has stated that the gender gap in Ethiopia’s agricultural productivity is 12%, and that closing the gap in the upstream segments of FVCs alone could lift more than 1.3 million people out of poverty. Addressing gender inequalities in all segments of FVCs in Ethiopia can lead to even more positive outcomes society-wide (Dethier & Effenberger 2011: 6-11, 38, 44-45; Gómez & Ricketts 2013; Masamha et al. 2018: 28-29).

(13)

1.1.2 Food Value Chains, Development and Women

While it is widely known that female stakeholders in FVCs are crucial for development, their labour is often not taken into account in policies that aim to exacerbate these developments. FVC development is understood as being a potential tool to fight malnutrition and health issues, for poverty reduction, economic growth, industrialisation, food security and sustainable food systems (Dethier & Effenberger 2011: 6-11, 38, 44-45; Gómez & Ricketts 2013; Masamha et al. 2018: 28-29). Next to this, the FAO (2013: 1) argues that “gender equality is central to […] achiev[ing] food security for all by raising levels of nutrition, improving agricultural productivity and natural resource management, and improving the lives of rural populations”. Productivity increase of FVCs is said to be especially necessary in Sub-Saharan Africa (SSA) in order for these countries to catch up with the rest of the world. Since the geographical agricultural land capacity is exhausted, meaning that we are running out of fertile agricultural land, the focus of FVC developments has to be institutional and economic in the hope to thereby increase productivity (Dethier & Effenberger 2011: 3). Institutional developments relate to market failures, missing markets and property rights. Economic developments include the investments and cooperation of governments, the private sector, donors, NGOs and civil society organisations (ibid.: 45).

Gómez et al. (2011: 1154) have argued that in order to adequately develop FVCs, all activities that are part of the FVCs have to be taken into account. A FVC involves all activities that take place from production till consumption in order to add value to a certain product, including growers, workers and consumers (ibid.: 1154; MAFF 2017: 2). FVCs consist of three segments: upstream, midstream and downstream. The upstream segment includes the sectors where a certain product is produced. The midstream segments consist of processing, storing and marketing. The downstream segments include distribution and consumption (Gómez et al. 2011: 1154; Minten et al. 2013a: 632-638).

The importance of the role of female stakeholders in developing (agricultural) FVCs has been recognised for decades, since women often are the primary producers of food in local settings. Despite the acknowledgement of the importance of inclusion of women, women’s labour contribution to FVCs has largely been invisible in economic and agricultural policy (Dethier & Effenberger 2011: 13; Mullaney 2012: 101). Especially women in the midstream segments have been neglected in development policies, despite their increasing importance in food systems as a result of structural changes in FVCs due to e.g. urbanization (Reardon 2015: 46). Economic

(14)

and agricultural policies focused on FVCs have been criticised for only benefitting large-scale farmers; overlooking certain relevant stakeholders through a limited focus on increasing productivity, which often only relates to the labour of men and ignores women’s labour; contributing to existing inequalities; and limiting the potential outcomes of FVC developments (Allwood 2013: 111; Dethier & Effenberger 2011: 13; Mullaney 2012: 101).

In order to overcome the aforementioned issues in FVC development policies, gender mainstreaming in policies has been proposed. Gender mainstreaming can be understood as to “systematically examine and address women’s as well as men’s needs, priorities and experiences as part of the development of policies, normative standards, programmes, projects and knowledge building activities” (FAO 2013: 9). It is argued that when data on women’s labour is infused into mainstream policies, projects and institutions, gender equality can be attained (Keleher 2012: 111). However, gender mainstreaming has often not led to the expected outcomes (Bosma et al. 2018: 1-2; Keleher 2012: 116; Okali 2012: 4). A problem of gender mainstreaming is that it can function as a so-called ‘tick-box’ system for policymaking, without resulting into increased gender equality and better gender relations ‘on the ground’ (Allwood 2013: 43-44; Mullaney 2012: 102). Another known problem in development policies is the portrayal of women as a homogeneous category and as outside the social and political context (Okali 2012: 2). Therefore, it is necessary that the local context of gender relations is understood and taken into account in policymaking (Mullaney 2012: 102). This research aimed to do so by studying the lives of the female injera bakers in their local context.

1.1.3 The Teff Value Chain and the Role of Women

This research explores the case of the teff value chain in Ethiopia, in which the failure in FVC developments to incorporate the female stakeholders and the neglect of women’s labour in Ethiopia more generally come together. Teff (Eragrostis tef) is the edible seed of the grass with the same name and has its origin in Ethiopia and Eritrea. Teff is Ethiopia’s biggest cash crop and an estimated 6.3 million households depend on the production of the low risk crop for their income and/or subsistence (Ethiopian Agricultural Transformation Agency et al. 2013: 1-3; Minten et al. 2015: 268; Minten et al 2013a: 1). The literature on the teff value chain reveals that, in 2013, the share of women of the truck drivers, rural wholesalers, urban wholesalers and female-headed teff farming households were all approximately 5%, and that the share of female urban retailers was approximately 15% (Minten et al. 2013b: 3). The available literature on women midstream states that their contribution is limited (Minten et al. 2015: 270). Regarding

(15)

the share of female teff farmers there are no hard numbers. The literature numerically only estimates that 95% of famers’ households are male-headed. However, it is known that women execute 65% of the selection and sorting of the teff seeds before they are sawn (Tegene et al. 2016: 49). All in all, the literature on teff suggests a binary and gendered gap with women being mainly included in the upstream segments of the chain, while men control the higher-valued commercial end: the downstream segments.

The Ethiopian government is interested in increasing the productivity and profitability of the

teff value chain in order to make it a source for poverty reduction and improved food security

(Ethiopian Agricultural Transformation Agency et al. 2013: 1-3; Minten et al. 2015: 268; Minten et al 2013a: 1). Consequently, policy recommendations try to increase the understanding of how the teff “food value chain function[s] and […] how opportunities can be harnessed from these changes to allow for better food security for rural as well as urban poor” (Minten et al. 2015: 275, 292). However, concerning the strategy on strengthening Ethiopia’s

teff value chain, the focus on human capital is limited to teff researchers, farm-owners, and

rural- and urban retailers: sectors dominated by men (Ethiopian Agricultural Transformation Agency et al. 2013). This limited focus thereby fails to include the potential of teff’s changing value chain to improve the socioeconomic situation of its female stakeholders, and vice versa.

Recently there have been many changes at the end of the teff value chain in Addis Ababa, the capital city of Ethiopia. Addis Ababa is quickly urbanizing and has seen a strong annual economic growth rate, which makes for the expansion of the food-service sector related to teff (Vandercasteelen 2018a: 383; Vandercasteelen 2018b: 393). The changes in the teff value chain, and in the urban food-service sector more generally, could have a big effect on the female

injera bakers in Addis Ababa that have been neglected by the formal institutions that try to

improve the productivity and profitability of the teff value chain (Minten et al. 2013a: 14). This research focused on the effects of the changes in the teff value chain on the visibility and possibilities for empowerment of the female injera bakers in Addis Ababa, Ethiopia.

1.2 Research Purpose

Much of the labour women undertake is made invisible due to gendered assumptions about the economy and labour more specifically. As a result, women’s labour is not well represented in policies that affect their situation. Developmental policies in FVCs perpetuate this invisibility

(16)

as well. While it is known that women play a crucial role in developing FVCs, and help to attain the expected outcomes, their labour is still largely excluded from such development policies. The Ethiopian government has articulated that it is willing to make women’s lives visible and to integrate their economic power and activities into economic planning. However, this is not reflected in most governmental policies and statistics, or in agricultural and FVCs development policies more specifically. Ethiopia is one of the SSA-countries that requires large FVC developments in order to catch up with other countries in terms of development. The policies on increasing the productivity of the teff value chain solely focus on ‘male’ labour and overlook the different ways women contribute to the chain. Recently, there have been many developments in the teff value chain that can have big effects on its stakeholders, including the neglected female injera bakers in Addis Ababa.

This research has three purposes. First, to explore the changes in the teff value chain over the last five years. Second, to examine the effects of the changes in the teff value chain on the visibility of the female injera bakers in Addis Ababa. Third, to study the effects of the changes in the teff value chain on the possibilities for empowerment of the female injera bakers in Addis Ababa.

1.2.1 Main Research Question and Sub Questions

This research will be led by the main research question: ‘How have the changes in the food value chain of teff affected the visibility and empowerment of female injera bakers in Addis Ababa, Ethiopia?’. This question is divided in three sub questions:

1. How have the changes in the teff value chain affected the injera food-service sector in Addis Ababa over the last five years?

2. How have the changes in the teff value chain affected the visibility of the labour carried out by female injera bakers in Addis Ababa?

3. How have the changes in the teff value chain affected empowerment opportunities for female injera bakers in Addis Ababa?

(17)

1.3 Structure of the Thesis

This thesis is made up of seven chapters. Chapter two outlines the key theories and debates surrounding invisible work, (in)visibility and empowerment, and addresses how these theories have informed this research. Chapter three describes the methods that have been used to define the unit of analysis, to collect and analyse the data, and the ethical considerations made. Chapter four provides the empirical context of this research. This includes a description of the specific research locations, Ethiopia’s socio-economic development indicators, a discussion of Ethiopia’s agricultural development policies, and how the changing teff value chain has affected the injera food-service sector in Addis Ababa over the last five years. Chapter five explains how the changes in the teff value chain have affected the visibility of the female injera bakers in Addis Ababa. Chapter six explores the effects of the changes in the teff value chain on the empowerment opportunities for the female injera bakers in Addis Ababa. Chapter seven provides the conclusions of the research discussed in this thesis, and discusses a multiple policy recommendations coming forth from the conclusion.

(18)

2. THEORETICAL FRAMEWORK

This chapter discusses the theoretical debates of the key concepts that have emerged from the research problem and how these concepts can be understood within the frame of this research. First, this chapter outlines a conceptual understanding of the changing teff value chain (see 2.1). Second, this chapter shines light on the inter-related concepts of invisible work and (in)visibility, and the difference between these two concepts (see 2.2). Third, it explains the concept of empowerment and the different theories on how to measure this abstract concept (see 2.3). Fourth, it addresses how these concepts are related within the frame of this research with use of a conceptual scheme (see 2.4). Lastly, it shines light on how the abstract concepts have been turned into measurable indicators with use of an operationalisation table (see 2.5).

2.1 Changing Teff Value Chain

One of the key concepts of this research is ‘the changing teff value chain’, which is hard to conceptualise since it includes many perspectives: geographical, economic, cultural, social, environmental, etc. This research does not focus on such a broad perspective. First, the focus will be on the changes which are taking place in the value chain, the characteristics of these changes (trajectories/factors/actors), and how these changes have affected the female injera bakers in the food-service sector in Addis Ababa. In short: it looks at the structuring conditions in the chain and how they influence the female stakeholders in the teff value chain in Addis Ababa, and vice versa. These conditions can be either enabling or constraining. Giddens (in Eriksen 1995 [2010]: 92) has defined such a relational construction as the duality of structure: through their agency, women ‘stakeholders’ shape the nature and structure of the value chain, which in turn affects their agency. It is a mutually constitutive and cautiously changing social process. By looking at the changing teff value chain as a duality of structure, it is recognised that individuals and society can only be conceptualised in interaction.

Second, the focus will be on how the changes in the teff value chain have affected the female

injera bakers by the increase in demand for injera. As became evident from the strategy on

strengthening the teff value chain, the female injera bakers’ labour is largely invisible. Therefore, the gap in knowledge and the critical issue to explore is the (in)visibility (see 2.2) of the female injera bakers now that the teff value chain has been changing.

(19)

Lastly, the focus will be on the ‘so what?’: what do these changes actually mean for the women themselves and their personal lives? The answer to this question will be led by the normative definitions of empowerment (see 2.3). The analysis focuses on the last five years, since the strategy on strengthening the teff value chain was published approximately five years ago.

2.2 Invisible Work/(In)Visibility

2.2.1 Invisible Work

The invisibilisation of women’s labour goes beyond the policies on increasing the productivity of the teff value chain and appears throughout the global economic discourse. The neoliberal conceptualisation of the economy assumes women’s labour to belong to the family- and household economy and assumes that men’s labour fits the market economy (Bryson 1996: 208). This binary gendered definition has become embedded and reproduced throughout the economy through institutions, processes, policies, etc., and have resulted in the institutionalisation of gendered inequalities throughout the global economic discourse (Pouw 2017: 7). These gendered inequalities include, among other things, that women earn less than men for the same labour, that women are less likely to be hired than men, and that women’s labour is often not perceived as labour at all.

The neglect of the labour executed by women in the economic discourse is also understood as ‘invisible work’. In 1987, Arlene K. Daniels coined the term ‘invisible work’ to “describe those types of women’s unpaid labour – particularly housework and volunteer work – which were culturally and economically devaluated” (Hatton 2017: 336). Today, the concept has been extended to include paid labour as well. Int his definition, ‘invisible work’ can comprehend ignored, overlooked, socially marginalised, economically devalued, culturally devalued, legally unprotected and/or unregulated labour (ibid.: 337). In turn, all these variables fall under three dimensions: spatial, legal and cultural. Spatial labour invisibilities occur “when labour is devalued because it is physically segregated from a culturally defined worksite” (ibid.: 337), legal labour invisibility occurs “when labour is devalued because it is excluded from legal definitions of ‘employment” (ibid.: 337) and cultural labour invisibility occurs “when labour is devalued by virtue of hegemonic cultural ideologies” (ibid.: 337). The latter is not limited to gender alone but can intersect with other social constructs such as race, class, gender, age and ability. Labour can thus be ‘multiply invisible’ as well (ibid.: 345).

(20)

2.2.2 (In)Visibility

Upholding the invisibility of women through disregarding and/or overlooking their participation and hence excluding them from policies that aim to increase their empowerment could also be seen as keeping these women subaltern. Spivak (2003: 25) has described the subaltern as a person living on the margins of society, without agency or social status. Subaltern differ from the oppressed because the oppressed “can speak and know their conditions” (ibid.: 25), while the subaltern do not have voice/resources and/or consciousness/awareness of their situation (ibid.: 27): they are silenced through the ideological construction of gender that keeps the female unrelated to the economy and their labour invisible.

Kabeer et al. (2013: 250) have argued that when labour is denied social recognition women often embody this denial. It could be that the female injera bakers also do not acknowledge their own labour as valuable or as labour at all. Therefore, keeping these women and their labour invisible is a silencing mechanism: they cannot self-organise to better their socio-economic position unless they are made conscious of their position as labourers (ibid.). Thus, as long as policies only refer to the ‘male’ labour, women’s (economic) empowerment is hard to be achieved through such policies, and can only be done when the women are conscious about their situation and in the position to express and voice their opinions.

The difference between invisible work and (in)visibility as conceptualised in this chapter lies in the perspective of the invisibility. Invisible work includes a constructed invisibility by third parties such as the government or the society more generally, but not by the workers themselves. (In)Visibility relates to how these women see themselves, to what extend they have embodied their invisible status, to what extend they are conscious about their social position, and to what extend they are able to voice their opinions about their situation. In this sense, invisible work is the embodied structure within which these women and their work are constructed and (in)visibility applies to the agency that these women execute in relation to their social position.

2.3 Empowerment

The term ‘empowerment’ has been applied since the 1970s, first mainly in feminist movements, Freudian psychology, the Black Power Movement and Ghandism. In the 1990s, the concept of empowerment entered the discourse of gender and development, thus concerning women’s empowerment and promoting gender equality within the development framework. The term

(21)

‘empowerment of women’ was endorsed at the World Conference on Women in 1995 and referred to the transformation of gender relations and other power relations with which women are linked (Drydyk 2013: 250). The Gender Empowerment Measure (GEM) conceptualises gender empowerment as the ability of women to actively participate in social and economic life (ibid.: 249). Amartya Sen (1999: 203) has defined empowerment as the expansion of agency, and agency as what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important. In order to achieve this, Sen acknowledges the importance of access to institutions, especially democratic ones, civil liberties and free press (ibid.: 297). Empowerment may then result from people’s larger agentive capabilities through democratization. Sen’s capabilities approach has been criticized for its neglect of how institutionalized power can create the opposite of the desired effect as well, and thereby can enhance inequalities (Hill 2003: 117). Conceptualisations of empowerment like Sen’s are often said to simply refer to ‘agency’ instead of ‘empowerment’, since they do not reflect the weight of the expanded agency over time (Drydyk 2013: 253).

Naila Kabeer (1999: 436) also refers to empowerment as a process of change in which the less powerful increase their ability to make choices. In Kabeer’s definition, the notion of choice is dependent on three inter-related dimensions: resources, agency and achievements. Resources include the material, human and social means that can enhance the ability to exercise choice. This dimension entails the distribution of allocative and authoritative resources as a consequence of society’s norms and values, and thus reflect the conditions for empowerment (ibid.: 437). Agency in relation to power can go two ways: it either reflects the capacity of an actor to define and pursue their own choices, or it reflects the capacity of an actor to override another actors’ agency (ibid.: 438). Kabeer then goes further in her explanation of empowerment by including achievements: she argues that the outcomes, if they reflect a constraint on the ability to make choices, are also an issue of power and should therefore also be taken into account when measuring empowerment (ibid.: 438). It is argued that the power to make decisions with ‘good’ outcomes is enhanced when people with little power are engaged not only with the more powerful, but also with each other, since collectives have more power than individuals (Drydyk 2013: 258). Pereznieto & Taylor (2014: 236) have defined the change outcomes further by outlining four dimensions of power:

1. Power within: the knowledge, individual capabilities, sense of entitlement, self- esteem, and self-belief to make changes in their lives, including learning skills to get a

(22)

job or start an enterprise.

2. Power to: economic decision-making power within their household, community, and local economy (including markets), not just in areas that are traditionally regarded as women’s realm, but extending to areas that are traditionally regarded as men’s realm.

3. Power over: access to and control over financial, physical, and knowledge-based assets, including access to employment and income-generation activities. 4. Power with: the ability to organise with others to enhance economic activity and rights. (Pereznieto & Taylor 2014: 236)

Within this research, I focus on Kabeer’s conceptualisation of empowerment but replace her conceptualisation of ‘good outcomes’ with the change outcomes and its four dimensions of power in order to come to a more measurable conceptualisation.

2.4 Conceptual Scheme

The conceptual scheme (see figure 1) illustrates the relation between the concepts that have been discussed in this chapter: the changing teff value chain, invisible work, (in)visibility and empowerment. (In)Visibility and invisible work are illustrated as overlapping concepts, because both aim to conceptualise the same phenomenon, but from a different perspective (structure vs. agency). As illustrated with the arrows, all concepts are interrelated, meaning that all can be affected by- as well as having an effect on the other concepts.

(23)

2.5 Operationalisation

(24)

3. METHODOLOGY

This chapter explains the research design of this thesis, the choices made in the research design, and its limitations. This chapter opens with the ontological and epistemological stance that underpin this research (see 3.1). This is followed by an explanation of the choices made concerning the unit of analysis and population sampling (see 3.2), the data collection methods (see 3.3) and the data-analysis methods (see 3.4). Next, this chapter evaluates the quality of this research (see 3.5) and the ethical considerations made (see 3.6). This chapter ends with a personal reflection on my influence on the research (see 3.7).

3.1 Ontological and Epistemological Stance

This research was approached from a constructivist ontology. This is the belief that social entities should be considered “social constructions built from the perceptions and actions of social actors” (ibid.: 32). Value-chains are an element of structure and injera bakers position themselves between their own agency and the constraints of existing value-chain structures and government policy. In line with this ontological stance, the epistemology approach was interpretivist. This stance acknowledges that “social reality has a meaning for human beings and therefore human action is meaningful” (Bryman 2012: 30) and therefore aims to interpret the actors’ actions and their social world from their point of view (ibid.: 30).

These stances were adopted into this research by focusing the interviews on how the female

injera bakers reflect on their own visibility and empowerment, and thereby to let them give

meaning to their own situation. An interpretivist stance also believes that there are multiple meanings given to the same situation, because reality is socially constructed differently by different actors. Therefore, I decided to incorporate the perspective of the injera retailers in Addis Ababa in the research problem as well, which can reflect the structural constraints of the

injera bakers. Adopting these ontological and epistemological stances allowed the research to

be embedded in the socio-cultural understandings of the subject matter.

3.2 Unit of Analysis and Population Sampling

This research consists of two groups of actors: (1) the female injera bakers and (2) the injera retailers in Addis Ababa. I decided to focus on the female injera bakers in Addis Ababa since they are the main focus in the above-stated research problem. The injera retailers in Addis

(25)

Ababa are the traders that sell the injera to the clients directly, among which the managers of the formal injera enterprises and the owners and employees of shops that sell injera. They were adopted into the research to reflect on the different aspects of the research problem, and to incorporate their perspective of the female injera bakers.

Concerning the female injera bakers, it is unknown how many (female) injera bakers are present in Addis Ababa. Therefore, there was no sampling frame available for this respondent group. Hence, I decided to use the method of ‘convenience sampling’, which means that the sample of female injera bakers was based on accessibility within the financial limits and time-frame of this research. For the formal sector, ten female injera bakers were located in four different enterprises. I found one out of the four enterprises online. The other three enterprises and the ten female injera bakers in the informal sector were located by my research assistant through asking her family, friends and neighbours, and through snowball sampling. My research assistant selected most respondents, as she already knew many people in the sector and had easier access to the respondents than me. The selection of the bakers was based on gender, only female respondents were included, because (injera) baking is predominantly a women’s job in Addis Ababa and Ethiopia more generally, and this research provides insight into the often overlooked ‘female’ sectors in the teff value chain. However, we also did not come across any male injera bakers during our search for respondents.

The injera retailers consisted of the managers of the formal injera enterprises and the owners and employees of shops that sell injera. The shops that sell injera were located in the public space and were therefore easy to access. The shops and respondents were found by walking through randomly selected areas of Addis Ababa and approaching every balitena1_{, bread shop,}

restaurant, supermarket, or souk2_{where they were selling ready-to-eat injera. The sampling of}

these respondents was done randomly (random sampling method), and the shops and respondents were found in five different sub-cities3_{(Addis Ketema, Kirkos, Kolfe Keranio,}

Lideta and Yeka) and 12 different woreda4_{. For the injera retailers in Addis Ababa there was}

1_{A balitena is a small shop where different dried spices, dried injera and sometimes ready-to-eat injera are} being sold.

2_{A souk is a small shop where household products, shelf life products, cigarettes, liquor, and sometimes} ready-to-eat injera are being sold.

3_{‘Sub-cities’ is a term used by the Ethiopian government to describe the different boroughs of the city.} 4_{‘Woreda’ is the Ethiopian administrative division below the sub-cities and can be understood as the different} neighbourhoods within a sub-city.

(26)

also no sampling frame available. However, by selecting the respondents as well as the shop-locations randomly throughout the city, I aimed to get a frame that still relatively accurately represents the injera food-service sector in Addis Ababa.

3.3 Data Collection Methods

For this research I used a mixed methods approach, meaning both quantitative and qualitative methods. Specifically, I used sequential mixed methods, in which the qualitative data-gathering was used as a principal tool, and the surveys were used to support and emphasize the qualitative dimension. Therefore, the qualitative data was gathered before the quantitative data. This decision was made because the emphasis of this research is on the female injera bakers’ reality primarily, and the perspective of the urban retailers secondarily. Consequently, the findings were triangulated, which entailed comparing the findings deriving from the interviews with the bakers in the formal sector, the interviews with the bakers in the informal sector, and the surveys performed with the urban injera retailers. Through triangulation the analysis provided a more valid and complete understanding of the research problem. The tools used to collect the data were semi-structured interviews and surveys, as well as participant observation (see table 2).

No. of respondents Male Female Location Data collection Informal

sector injera bakers

10 0 10 10 different homes Semi-structured interviews Formal sector injera bakers 10 0 10 4 different enterprises Semi-structured interviews Managers injera enterprises 3 3 3 enterprises Semi-structured interviews Urban injera retailers 40 13 27 6 bread shops, 11 balitena, 19 souks and 3 supermarkets Surveys Total 63 16 47 - 23 semi-structured interviews, 40 surveys

(27)

3.3.1 Semi-Structured Interviews

The decision to use semi-structured interviews was made because of two reasons. First, I wanted to get rich and detailed answers from the respondents, because I wanted them to explain their constructed reality for me, hence structured interviews would be too limiting. Second, doing unstructured interviews would pose a risk of having faulty translations and interpretations, because the interviews were conducted in a different language than the original interview guides. By conducting semi-structured interviews, the interview guide (see appendix I, II and III) could be translated to Amharic by the research assistant before-hand. This way, I could check the appropriateness of the interview with my research assistant, we could check whether the research assistant and I interpreted the questions the same way and whether they were translated correctly, while still keeping the possibility of having rich and detailed answers. For the same reason it was decided to translate the interviews at ‘real-time’, as well as to give way for me to jump in when there were answers that needed more specification or explanation.

The semi-structured interviews were developed starting with ‘face sheet’ information, and from there going into the different topics as defined in the operationalisation table (see table 1). I chose to have mostly open-ended questions, in order to stimulate detailed answers to the questions. Still, some of the questions were closed-ended. The interviews with the informal female injera bakers were an improved version of the interviews with the formal female injera bakers: the latter included more questions per topic in order to collect more complete information per theme. It did not include any questions on different topics than those addressed in the interviews with the injera bakers in the formal sector.

All interviews took place at the site of the enterprises. I chose to do so because it meant easy access to the respondents and it took little time out of the respondents’ day. Also, doing the interviews somewhere else was difficult since Addis Ababa is hard to navigate because there are almost no street names and public quiet places are very uncommon. As a result, the sites at which the interviews were held were sometimes disturbed by noises of machines, customers walking in, employees asking questions, or even children running around. In some instances, the employers of the female injera bakers would come in during their interviews, and although the questions were more about their personal lives than about the enterprise, it is inevitable that this has influenced the information given. Most probably, the information had a more positive outlook, in order to seem satisfied about their job to their employer. The interviews with the informal female injera bakers took place at the homes of the respondents, which were

(28)

simultaneously their worksites. Some respondents were still baking injera, others had already finished baking. Sometimes there were family members around, but usually it was just the respondent, the research assistant and me. I chose to do the interviews in their homes, since they were all busy with baking, looking after their children or doing other household chores, and therefore did not have the time or opportunity to meet elsewhere.

3.3.2 Surveys

The surveys were developed starting with ‘face sheet’ information, and from there going into the different topics as defined in the operationalisation table (see table 1) and based on the previously written interview guides. The survey-outline (see appendix IV) was sent to the research assistant before-hand in order to check the cultural and social appropriateness of its contents and was thereafter translated to Amharic. Despite the translation, the surveys were filled out by the research assistant. This ensured that the surveys were understood correctly and was done to evade asking the respondents whether they were literate and thereby to pressure those respondents that were not. The forty surveys were conducted in four days. The research assistant conducted ten surveys by herself after we had already done two days of conducting surveys. She was given instructions on how to conduct the surveys, and the site where she conducted them was known to me. Since the surveys were conducted publicly, there was always the possibility of third persons to interrupt the administering of the surveys. Whenever there was a customer or another third person, we would stop the survey and return to it when they had left. Since the research assistant had checked the contents of the surveys for any inappropriateness, I decided not to pre-test the questionnaires in a pilot before proceeding to administer them fully.

3.3.3 Participant Observation

I did participant observations throughout the nine weeks that I was present at the research location, which included more observations than participation. The participant observation included going to markets (Mercato and Shola market), shops, restaurants, injera enterprises and people’s homes. Most of the participant observation was not done with any specific criteria or in a planned manner, since it involved my daily life in Addis Ababa. Injera is a very public good, since it is sold in shops on the streets and eaten by everyone usually three times per day. Furthermore, the conversations that were part of these observations came about naturally as a result of people’s interest in the reason for my presence in Addis Ababa and their eagerness to

(29)

tell me more about the subject and their culture more generally. While conducting the interviews and surveys, observations were made of the sites and the surrounding areas (see 4.1). In all conversations about my research subject during the observation my role as a researcher was overt, and the setting was usually closed. Most other observations were done with a covert role in open settings. All observations were written down in a notebook at latest in the evening of that same day. The notes included information on the date, the daypart, the location, the present actors and their characteristics.

3.4 Data Analysis Methods

The interviews were analysed using ATLAS.ti data management software. With use of ATLAS.ti the key thematic codes (see table 3) were identified using open coding at first, meaning an inductive approach was used leading to thirteen codes. Next, a deductive approach was applied to the data with eleven pre-set codes based on the theoretical framework. Consequently, all twenty-four codes were grouped into key themes. Hereafter, the coded data was analysed for patterns (frequency in word-use) in order to build the theory. The surveys were analysed using SPSS Statistics, using descriptive statistics and custom tables that identified correlations between the variables.

Key themes Open coding codes Pre-set codes

Empowerment Empowerment: achievement {40-0}

Empowerment: agency {38-0} Empowerment: resources {49-0} Children’s life as life goal {22-0}

Expenditures {42-0} Free time {25-0} Future {31-0} Independence {22-0} Physical state {34-0} Survive {38-0} Thankful {18-0}

Invisibility Invisibility: consciousness {138-0}

Invisibility: voice {94-0}

Changing teff value chain Nature and structure of the chain shaped by stakeholders {0-0} Conditions: enabling {28-0}

(30)

Conditions: restricting {18-0}

Invisible work Invisible work: cultural {23-0}

Invisible work: legal {21-0} Invisible work: spatial {21-0} Burden {42-0}

Double burden {14-0} Quality {25-0} Low threshold {12-0} Only option {16-0}

Table 3. Overview of codes and key themes.

3.5 Evaluating the Quality

The quality of the quantitative part of this research was assessed using the traditional criteria (external reliability, internal reliability, internal validity and external validity) as conceptualised by LeCompte and Goetz (cited by Bryman 2012: 390) and the quality of the qualitative part using the criteria of trustworthiness by Guba & Lincoln (ibid.: 390-393). They have identified four criteria to evaluate the quality or ‘trustworthiness’ of qualitative research: credibility, transferability, dependability and confirmability. I chose to use this criteria because, as in line with my ontological and epistemological stance, it is based on the belief that there is no one truth about the social world (ibid.: 390) and because qualitative research is inherently different from quantitative research and therefore needs different quality measures.

The first criterion is credibility, which equals internal validity (ibid.: 49). This criterion relates to whether the research findings accurately reflect the social reality. In order to ensure credibility and internal validity, I used mostly to open-ended questions for the interviews and included opposite stances about the same subject for the closed-ended questions in the interview and for the surveys. An advantage of having semi-structured interviews is that it reduces error because of variation in how the questions are asked (ibid.: 210). Through the above-stated process (see 3.3.1 and 3.3.2) possible errors in translation and variation were minimised. The survey respondents were all obtained through random sampling, which increases the chances of the frame being truly representative of the bigger population (ibid.: 195). However, in a city of over 3 million people, the sampling frame of a total of 63 respondents was still very minimal. Guba and Lincoln argue that respondent validation and triangulation are best to ensure internal validity (ibid.: 390). Respondent validation was not part of the initial research design and can

(31)

hence be seen as a limitation. However, triangulation was used as a technique to cross-check findings and to ensure validity.

The second criterion is transferability, which parallels external validity (ibid.: 49). Transferability differs quite a lot from external validity, in the sense that external validity focuses on the breadth of the research and transferability focuses on the depth of the research. External validity focuses on whether the findings hold in another context, and transferability is about the richness of the findings (ibid.: 392). The specificity in time and place of this research means that transferability and external validity are not applicable, meaning that the findings do not hold if the study is replicated in another scene and that the results cannot be generalized. Also, the sample for the female injera bakers in Addis Ababa was acquired through convenience sampling. One of the general limitations of this method is that one cannot know for which population the responding sample is representative and this can pose a problem with regard to generalisation (ibid.: 201). The sample for the injera retailers was acquired through random sampling. Usually, this method of sampling is praised because the data gathered can be easily generalized (ibid.: 195). However, since no sampling frame of this research population was available and the research is based in a specific context, this assumption does not hold for this research. External reliability highly relates to the aforementioned criteria: it is the extent to which a study can be replicated. This is impossible in this research due to the specificity of the social setting. Generalisation and replication were never a goal for this research.

The third criterion is dependability, which equals internal reliability (ibid.: 49). Internal reliability refers to whether the members of a research team agree about the content of the research. Dependability can be ensured by using an auditing approach which includes keeping detailed records of all phases of the research process and then have auditors check the properness of the procedures. Auditing was not applied in this research. Regarding the internal validity, the research assistant and I discussed the interview- and survey outlines beforehand to check whether we interpreted (the context of) the questions in the same way and whether they were translated correctly. It is inevitable that there still have been translation problems, due to restraint, out of respect, in my research assistant to tell me if something was not clear or correct.

The fourth criterion is confirmability, which equals objectivity (ibid.: 49). Full objectivity is never possible, but it should include that the researcher has not influenced the research by letting known his or her own personal beliefs or values. The interviews and surveys did not include

(32)

any leading statements, and often included contrasting statements on the same issue in order to minimize the possible influence of the researcher. However, it is likely that the respondents felt that I had certain personal values concerning the subject due to my overt identity being a white/Western woman.

3.6 Ethical Considerations

Throughout the process of this research great care was taken to ensure that the ethical standards of voluntary participation, informed consent, safety in participation, confidentiality and trust were being met. First, in regard to voluntary participation, any objection to participate was always accepted. This happened twice: once during the surveys, and once during the interviews in the informal sector. We thanked them for their time regardless and moved on. All participation was voluntary, and many respondents were enthusiastic that a foreigner was interested in their culture and that I recognized the importance of the subject. Except for one instance where one of the respondents did not want to give his name, none of the respondents refused to answer a specific question.

Second, before both the interviews and the surveys the respondents were informed on the aim of the research and were asked for consent. All participants gave consent to do the interviews and to have them recorded. Consent was asked and given verbally, since written consent could have caused discomfort or have been impossible due to illiteracy.

Third, the safety in participation was more difficult to attain, since the interviews took place at the respondents’ worksites. This could have had safety implications, including their bosses or colleagues having a negative opinion about what was said. Still, I chose to do the interviews at their worksites, since most did not have the time or opportunity to meet elsewhere, and because it meant easy access to multiple respondents at once. This was a big compromise I had to make. On an emotional level, it was made sure that the respondents had the time to express their feelings about certain topics, and the last question of the interviews and surveys gave way to express anything else which they wanted to get off their chest.

Fourth, the personal information about the respondents was kept private from anyone besides me and the research assistant, and all respondents are stated anonymously throughout this thesis in order to ensure confidentiality. The personal information includes anything that can give

(33)

away the respondents’ identity, such as the names of the companies at which they work at and their own names. Quotes used in this thesis will be referenced with an interview number, but the interview number will not be linked to any ‘face sheet’ information.

Fifth, trust was one of the most difficult ethical standard to ensure, because building rapport was not possible. At some instances, respondents were suspicious of the fact whether I worked for the government. We could eliminate these suspicions by explaining the intent of the research again, and to show the questions we were going to ask. Still, it is likely that some respondents have held back information due to mistrust as a result of a lack of rapport.

3.7 Personal Reflection

My positionality as a white young lady in Ethiopia’s capital city has inevitably had an enormous influence on the research. First of all, being a woman has been both an advantage and a disadvantage. On the one hand, the female respondents could have felt that they could be more open in their experiences as being a woman. On the other hand, especially with male respondents, I was not always taken seriously immediately, or they felt the need to ‘mansplain’ things to me, which however still led to valuable information. Being white meant that I was always seen as an outsider, or in the case of Ethiopia as a ‘farenji’5_{. This often helped me,}

because people wanted to explain their culture to me and were happy that I showed interest in this. On the other hand, while no one expressed any negative assumptions about me to my face, it is obvious that people in a country that has seen a lot of Western interference will not always have a positive attitude toward me as a white woman going to Ethiopia for my own benefit. Concerning this issue, my nationality also did not help, since two Dutch people possessed a patent on teff in Europe which was ‘stolen’ from Ethiopia. Many respondents and other people I met in Ethiopia spoke to me about this, and I always distanced myself from these Dutch people. Surely there have been instances where respondents felt this way but did not express their feelings to me. This could have had a negative effect on the information they gave me. Concerning my own biases, assumptions and personal struggles, I found especially two things difficult. First was the extreme economic inequality. Some of the respondents were quite well-off, while others lived in rough circumstances. Knowing that I could not better their lives, but

(34)

still asking things from them often felt immoral. Second was the extreme patriarchy in Ethiopian society, which seeped through everything and seemed internalised by many of the people I met. This made me feel frustrated and sometimes pessimistic about my hopes for (greater) gender equality in Ethiopia in the near future. While experiencing this was difficult, I feel that it also made my research stronger, since it made me increasingly aware of the (contextual) importance of the topics addressed in this thesis.

(35)

4. EMPIRICAL CONTEXT

This chapter provides an overview of the empirical context in order to further emphasize the relevance of this research and to examine how the changing teff value chain has affected the

injera food-service sector in Addis Ababa over the last five years. First, this chapter describes

the specific sites where the interviews and surveys were conducted (see 4.1). Second, this chapter provides an overview of the socio-economic development indicators such as poverty (see 4.2.1), education (see 4.2.2), health (see 4.2.3), and gender equality (see 4.2.4). The chapter follows with a discussion on the agricultural development focus in Ethiopia and its related policies (see 4.3). Lastly, the empirical context of the changing teff value chain and its effect on the injera food-service sector in Addis Ababa over the last five years are discussed (see 4.4).

4.1 Research Location

Ethiopia is located in the Horn of Africa, and Addis Ababa lies right in the middle of the country (see figure 2). Ethiopia has an estimated population of 108 million people, of which approximately 17% live in urban settings. Addis Ababa has an estimated population of at least three million people6_{. The following descriptions are a mix of a general description of the}

sub-cities, as well as more detailed information of the specific sites the surveys and interviews were

conducted. All the descriptions below are based on (participant) observations.

The respondents for the surveys were found in five different sub-cities in Addis Ababa: Addis Ketema, Kirkos, Kolfe Keranio, Lideta and Yeka (see figure 3). Addis Ketema is the area of the vibrant Mercato: the biggest open-air market of Africa. At the location where the surveys were conducted there were many people walking around, men carrying baskets with all sorts of products, donkeys carrying big bags and many small shops (see figure 4). Four surveys were conducted in Addis Ketema, including one bread shop and three souks.

(36)

Figure 2. Map of Ethiopia Figure 3. Map of Addis Ababa by sub-city (source: agricaguide.com/country/Ethiopia). (source: http://www.addisababa.gov.et/es

web/guest/city-map).

Figure 4. A street near Mercato with a man carrying buckets, shops and other people walking by (source: own picture).

(37)

Kirkos is located at the heart of Addis Ababa and holds some of the biggest hotels in the city. It is one of the oldest areas in Addis Ababa, and therefore contains both old ‘villages’7_{that look}

largely self-made (see figure 5), as well as the more luxurious condominiums for high middle-class citizens. The inhabitants in this area portray a great mix of people. The shops in this

sub-city where the surveys were conducted reflect this mix, including both more luxurious shops

that looked very professional and smaller shops that did not sell as many products. Five surveys were conducted in Kirkos, including two balitena, two souks, and one supermarket.

Figure 5. A passage with souks in one of the old ‘villages’ of Kirkos (source: own picture).

Yeka (see figure 6) is one of the biggest sub-cities in Addis Ababa. The site where the surveys were conducted was close to a big condominium, in which usually higher middle-class people live. All shops looked well-organised and neat, and attracted both the inhabitants from the condominiums as well as people living further away in village-areas. In this area there are many beggars, including children, and street vendors selling books, CDs, nuts and more. Five surveys were conducted in Yeka, including one balitena, two souks, and two supermarkets.

(38)

Figure 6. One of the main roads in Yeka (source: own picture).

Lideta is situated next to one of the biggest hubs in Addis Ababa named Mexico. However, except from the main road where there are many large buildings with business offices, the area is comprised of village areas with brick roads and only small convenience shops (see figure 7). Sixteen surveys were conducted in Lideta, including six balitena, three bread shops, one restaurant, and six souks.

Figure 7. A street running through one of the villages in Lideta with small shops and garbage on the street (source: own picture).

Kolfe Keranio is located on the outskirts of the city, in the mountainous area that lies around Addis Ababa, from where one can see the whole city. There are many new big houses, however surrounded by old self-build houses that look less modern. The area is very open and green, and especially the people riding wooden carriages pulled by donkeys give this area the feeling of a village. Ten surveys were conducted in Kolfe Keranio, including six souks, two bread shops, and two balitena.

Looking beyond the visible : how food value chain developments affect its neglected stakeholders : case study of the female Injera bakers in the Teff value chain in Addis Ababa, Ethiopia

LOOKING BEYOND THE VISIBLE: HOW

FOOD VALUE CHAIN DEVELOPMENTS

AFFECT ITS NEGLECTED STAKEHOLDERS

Case Study of the Female Injera Bakers in the Teff Value Chain in

Addis Ababa, Ethiopia

Diede Smith

International Development Studies

Looking Beyond the Visible: How Food Value Chain

Developments Affect Its Neglected Stakeholders

Case Study of the Female Injera Bakers in the

Teff Value Chain in Addis Ababa, Ethiopia

MSc International Development Studies 2018-19

Graduate School of Social Sciences

Diede Smith

10753044

July 8

, 2019

Word-count: 24995

Supervisor: Drs. Josh K. Maiyo

ABSTRACT

ACKNOWLEDGEMENTS

TABLE OF CONTENTS

LIST OF FIGURES AND TABLES

LIST OF ACRONYMS

1. INTRODUCTION

1.1 Problem Statement

1.2 Research Purpose

1.3 Structure of the Thesis

2. THEORETICAL FRAMEWORK

2.1 Changing Teff Value Chain

2.2 Invisible Work/(In)Visibility

2.3 Empowerment

2.4 Conceptual Scheme

2.5 Operationalisation

3. METHODOLOGY

3.1 Ontological and Epistemological Stance

3.2 Unit of Analysis and Population Sampling

3.3 Data Collection Methods

3.4 Data Analysis Methods

3.5 Evaluating the Quality

3.6 Ethical Considerations

3.7 Personal Reflection

4. EMPIRICAL CONTEXT

4.1 Research Location

_{, 2019}