Mapping power relations in Malawi
Investigating structures of failure for the
provision of sanitation in Lilongwe City
David Caspers
MSc Thesis International Development Studies
Mapping power relations in Malawi
Master of Science Thesis by
David Alexander Caspers
Supervisor
Dr. Michaela Hordijk
Examination committee
Dr. Michaela Hordijk (University of Amsterdam)
Cecilia Alda Vidal (UNESCO-IHE)
This research is done for the partial fulfilment of requirements for the Master of Science degree at the University of Amsterdam, Amsterdam, Netherlands.
San Sebastian
August 2016
Front Cover Photo: Family walking through treatment ponds at Kauma sewage treatment plant,
©2016 by David Alexander Caspers. All rights reserved. No part of this publication or the information contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording or otherwise, without the prior permission of the author. Although the author and University of Amsterdam have made every effort to ensure that the information in this thesis was correct at press time, the author and University of Amsterdam do not assume and hereby disclaim any liability to any party for any loss, damage, or disruption caused by errors or omissions, whether such errors or omissions result from negligence, accident, or any other cause.
Abstract
Legally mandated actors and actual practices are constantly shaping the sanitation sector in countries of the global south; a mixture associated with power differences and conflicts. This Master Thesis fills the knowledge gap of differences between practiced and legally mandated responsibilities actors assume for the development of the sanitation sector in Lilongwe City, and the ways power relations between those actors shape both of these responsibilities. Gaventa’s power cube has guided qualitative data gathering using concepts of forms, levels and spaces of power for the analysis of power relations between actors involved. Quantitative data has been gathered through power mapping. For both the power cube and the power-mapping tool implications of findings for theory and practice are discussed.
Findings include that existing power structures in Lilongwe are defined through a broad diversity of factors, ranging from trends of a democratic culture of obedience and silence towards authority to an observable lack of financial and human resources. Lastly existing national institutional and legally binding decisions and legislation regarding sanitation provision are causing conflicts amongst actors involved. Several clashes on various levels of government hinder the sector from moving on, while negotiations for re-assignment of responsibilities are moving only slowly. The thesis concludes that solutions are at close hand, but require the strong political and practical will of only a handful of mainly national actors to closely work together. This thesis furthers the practical application of concepts for understanding power, whilst answering important and urgent questions regarding the assignment of responsibilities bottle-necking the development of the sanitation sector in Lilongwe from a perspective of power.
Keywords:
Power relations; Power analysis; Sanitation; Lilongwe City; Malawi; Government of Malawi; Power cube; Power mapping;
Acknowledgements
I would like to thank Michaela Hordijk for the supervision of my thesis and the valuable input given throughout the process. I would furthermore like to thank Maria Rusca for giving advice in the field, and thank Cecilia Vidal for offering me her advice and input. I also want to thank the whole research team of fellow students for being part of this research, and helping to bridge the time between interviews in the field with joy and fun. Lastly, I want to thank my family, friends and of course my dearest Nele, for supporting me wherever possible.
Table of Contents Abstract i Acknowledgements iii Table of Contents v Abbreviations viii 1 Introduction 1 1.1 Introduction 1
1.2 Problem statement and research objectives 3
1.3 Thesis structure 4
2 Theoretical framework 6
2.1 Introduction 6
2.2 Conceptualizing power 6
2.3 Power and access to sanitation in a development context 7
2.4 The Power Cube by Gaventa 8
2.4.1 Spaces of power 8
2.4.2 Forms of power 9
2.4.3 Levels of power 10
2.4.4 Interrelationship of spaces, places and forms of power 10
2.5 Summary 11
3 Methodology 12
3.1 Introduction 12
3.2 Research question 12
3.2.1 Sub-questions 13
3.3 Epistemic assumptions and implications for research design 14
3.4 Research design 14
3.5 Unit of analysis and unit of observation 14
3.6 Conceptual Framework 15
3.7 Operationalization table 17
3.8 Sampling strategy 18
3.9 Methods of data collection 18
3.9.1 Literature review 18
3.9.2 Power mapping game (PMG) 19
3.9.3 Semi-structured interviews with actors involved in sanitation development 22
3.10 Triangulation 22
3.11 Mode and methods of data analysis 22
3.12 Ethical considerations 23
3.14 Combining the PMG with the power cube 25 3.15 Summary 25 4 Context Setting 26 4.1 Introduction 26 4.2 National context 26 4.3 Historical developments 27
4.4 Actors involved in developing sanitation in Lilongwe City 28
4.5 Strategies for improvement and development of the sector 29
4.6 Main issues in the sector 31
4.7 Summary 33
5 Legal responsibilities of actors of sanitation development in Lilongwe City 34
5.1 Introduction 34
5.2 Legal responsibilities of actors 34
5.3 Regulatory framework 37
5.4 Legal conflicts bottle-necking the legal development 39
5.5 Summary 41
6 Practiced responsibilities of actors of sanitation development in Lilongwe City 42
6.1 Introduction 42
6.2 Practiced responsibilities of actors on the ground 42
6.2.1 The LCC and the LWB 42
6.2.2 National state actors 48
6.2.3 Non state actors 49
6.3 Comparison of legally mandated and practiced responsibilities 51
6.4 Summary 54
7 Power relations between actors and their effects on shaping responsibilities 56
7.1 Introduction 56
7.2 The 10 most important actors with power values 57
7.3 Sources of power 59
7.4 Most important actor groups 61
7.5 Collaboration of actors between levels of power 62
7.6 Forms, spaces and levels of power used in sector 63
7.6.1 Spaces of power present in sector 63
7.6.2 Forms of power present in sector 64
7.6.3 Levels of power present in sector 65
7.7 How power relations shape responsibilities in the sector 66
7.8 Summary 68
8 Discussion 69
8.1 Methodological reflection 69
8.2 Key variables influencing the responsibilities within the sanitation sector 70
8.3 Recommendations for policy and practice 72
Appendices 78
Appendix A – PMG (Schiffer, 2007, p. 9) 78
Appendix B – Power Matrix Design by Caspers (2016) 79
Appendix C – Interview Questions by Caspers (2016) 80
Appendix D – 20 top named actors involved in sanitation sector in Lilongwe City 81
Appendix E – All actors involved in sanitation in Lilongwe City 82
Appendix F – LCC Structure 83
Appendix F - Actor groups parent grouping 84
Abbreviations
ADB – African Development Bank
DfID – Department for International Development (Part of UKAid) DoES- Department of Engineering Services
DoHSWS – Department of Health and Social Welfare Services DP- Development Partner
EU- European Union
GoM – Government of Malawi
INGO – International non-governmental organisation JSR- Joint Sector Review
LCC - Lilongwe City Council LIA- Low Income Area LWB - Lilongwe Water Board
MoAIWD - Ministry of Agriculture, Irrigation and Water Development MoE- Ministry of Education
MoF - Ministry of Finance MoH- Ministry of Health
MoLG- Ministry of Local Government NGO – Non-governmental organisation ODF – Open Defecation Free
PMG – Power Mapping Game
SDG – Sustainable Development Goal SWAp – Sector Wide Approach TA - Traditional Authority WB- World Bank
1 Introduction
1.1 Introduction
Sustainable development requires the collective effort of actors on all levels where power is present, ranging from global to national to local levels. This also holds true for the achievement of the Sustainable Development Goals (SDG), more specifically SDG 6.2, which states that by 2030, adequate and equitable access to sanitation and hygiene for all shall be achieved, with paying special attention to the needs of women and girls, and those in vulnerable situations. Moreover, SDG 6.b determines the will to “support and strengthen the participation of local communities in improving water and sanitation management”(General Assembly, 2015). A case of exceptional relevance for sustainable development is the capital of one of the poorest countries in the world, namely Lilongwe City in Malawi. In this city, provision of water borne sanitation is far off from being widely available, with only 9% of the population of the city being connected to the sewerage system while 20% use septic tanks (JICA, 2010). Rapid urbanization is estimated to be happening with an annual increase of 6%1 regarding population within the city borders of Lilongwe, and resources for future developments are scarce within national and local government2. As the UN notes, rapid and unplanned urbanization can have negative effects on sustainable development when infrastructures and policies do not aim at an equitable distribution of benefits for the whole city (UN, 2014). Looking at sewage infrastructure in Lilongwe City, sustainable development is endangered, as the existing and limited infrastructure heavily benefits local elites only, with the poor being practically excluded from accessing the benefits in any useful way. Based on national reports, the proportion of people with access to basic sanitation increased from 72,4% to 95,1% in 2013 (Ministry of Finance, 2014)3, and increasingly more regions in Malawi are declared Open Defecation Free (ODF) (Government of Malawi, 2011). However, other reports declare that since 1990 only 25% of Malawi’s population has gained
1
Note from field book
2
Interviews D-5, D-13, D-26
3
The government of Malawi is working with different definitions of safe sanitation, reducing reliability of provided data.
access to improved sanitation, summing up to a total of a meagre 41% (Hunga, 2016)4. Having access to improved sanitation does not imply that one is connected to a sewer or a septic tank per se, but only implies that one uses a facility which separates human excreta from human contact (WHO & UNICEF, 2012) – and can thus also be only a pit latrine with slab. The Government of Malawi (GoM) explicitly recognizes that key bottlenecks are limited and declining financing for the sector, with a 40,8% drop in funding in the overall national budget from 2013 compared to the year before(Government of Malawi, 2014). This drop can be attributed to the Cash Gate Scandal, in which many high officials of the government of Malawi have been found pocketing more than 250 million USD, causing a major international donor crisis for the country (Gieler, 2016). Funding for the sanitation sector of Malawi has been less than 2.1% of the national budget over the last few years (WaterAid Malawi, 2010). Furthermore high levels of non-functionality of water supply and sanitation facilities, dwindling raw water resources and inadequate institutional and human capacity to effectively plan, design implement and support the sector’s priority services (Government of Malawi, 2014) are identified as bottlenecks. The case of Lilongwe City thereby provides a prime example of a challenging situation for all actors involved in helping to universally achieve the SDG’s, specifically with regards to sanitation and the inclusion of all inhabitants of the City of Lilongwe in a just and equitable manner.
This research has been undertaken in the framework of the collaboration between the
UNHIDE (Uncovering Hidden Dynamics in Slum Environments) and INHAbIT Cities (Investigating Natural, Historical and Institutional Transformations in Cities). UNHIDE is sponsored by the Netherlands Ministry of Development Cooperation (DGIS) and is
implemented under the MoU between University of Amsterdam and UNESCO-IHE. INHAbIT
Cities is sponsored by the European Union’s Horizon 2020 research and innovation program
under the Marie Skłodowska-Curie grant agreement No 656738 and is implemented by King's College London. This research is thereby supplementing the UNESCO UNHIDE project with data and results. UNHIDE corresponds to “Uncovering Hidden Dynamics in Slum Environments and focuses on the interaction of ecological, technological and social processes that shape the urban waterscape in the Global South”(UNESCO-IHE, 2015). The UNESCO UNHIDE Project has already mapped out the situation regarding the access to clean drinking water in Lilongwe,
4
GoM makes a difference between Basic Sanitation and Improved Sanitation (see DeGabriele 2009, p. 10)
Malawi. Acknowledging the interrelatedness of issues of water and sanitation, the UNHIDE Project has been still lacking research to be carried out regarding access to sanitation, which this research is contributing to as part of a larger collaborative effort with fellow Master Students from the University of Amsterdam.
1.2 Problem statement and research objectives
The study of power relations in any given geographically limited sanitation sector carries academic and societal value, for it informs researchers and members of society alike about the perceptions of power and power relations between actors. The influence of power relations on sectorial development processes has been confirmed by studies (D Satterthwaite, McGranahan, & Mitlin, 2005), especially within the domain of political ecology that examines political dynamics surrounding material and discursive struggles over the environment (Bryant, 1998). Recent studies are arguing in favour of equal, transparent and accountable relationships of power, as the marginality and vulnerability of the poor increases with unjust or colonially influenced power relations. Equitable distribution of benefits is required for sustainable development (UN, 2014), and not following policies supporting this goal can harm the ability of achieving the SDG’s, and moreover create social tensions within a city that can have additional negative side effects.
For Lilongwe City, a knowledge gap can be recognized specifically with regards to legally mandated and practiced responsibilities of actors, especially with regards to the power relations in place and the effects of these power relations on the assumption of responsibilities by actors involved. As mentioned earlier, flush toilets and water borne sanitation are still an exemption for most of the HH in Lilongwe City (only 9% being connected to the sewerage system) and the most prevailing type of toilets used are pit-latrines (Langkau, 2016). In light of the newly established SDG 6, and the UNESCO UNHIDE research project, this bears the following research question: How are legally mandated responsibilities and practiced responsibilities taken up by actors for the development of the sanitation sector in Lilongwe, Malawi differing from each other, and how are those responsibilities affected by power relations that actors are embedded in?
This thesis contributes to the development of the sanitations sector through informing its actors with this research about the currently existing power relations and power structures that govern the development of legally mandated and practiced responsibilities amongst actors. In
light of the UNESCO UNHIDE Project, this research contributes specifically through mapping actors, practiced and legally mandated responsibilities and power relations between actors involved at the international, national and local level. This thesis will analyse the forms, spaces and levels realms of power, thereby providing information and data to be followed up upon by others in order to create, design and ultimately implement mechanisms, strategies and policies for improvement. The research on governance and power structures that are influencing the development of the sanitation sector in Lilongwe City thereby manifests a high societal relevance. Gaventa rightly notes, that if we “want to change power relationships, e.g. to make them more inclusive, or just more pro-poor, we must understand more about where and how to engage”(Gaventa, 2006a, p. 23).
Methodologically this research further develops the concept of the power cube by Gaventa (2006) in its use for analysis of power relations between actors within a policy field. This is achieved through using the power cube in a power context it has not been initially designed for - namely a space not characterized by bottom up civic participation, but conversely rather determined by a governmental driven top down structure. Gaventa´s power cube will be explored in its applicability in this other, more governmental context, with required adaptions for theory and practice discussed. It thereby adds scholarship, knowledge and experiences to the existing literature regarding the power cube and its applicability in the global south. Lastly, the power-mapping game (PMG) as developed by Schiffer (2007) is applied in this research. It has been used in helping gather data about power relations and power-perceptions from actors in the field, firstly about themselves and secondly about the other actors involved in the arena of actors responsible for the development of the sanitation sector in Lilongwe City. While the PMG has been used in many comparable developmental settings for the mapping of actors, it has so far not been used in explicit combination with Gaventa’s power cube. This thesis thereby adds scholarship and methodological reflection of the usefulness of combining both approaches for research.
1.3 Thesis structure
The first chapter gave an introduction to the problem statement and its research objectives, and is herewith concluded by a preview of the remaining thesis structure. The structure of the remaining chapters of the thesis is as follows.
The second chapter is the theoretical chapter, which introduces the theoretical framework in which this research is grounded. This includes a conceptualization of power in general, power in the WASH sector specifically and the concept for analysis of power used, namely the power cube by Gaventa.
The third chapter elaborates upon the methodology being used for this research, introduces the research question and sub questions, elaborates on the research design, modes and methods of analysis and limitations. It also provides a methodological article on combining the power cube by Gaventa with the tool for power mapping as developed by Schiffer.
The fourth chapter provides the required context setting, elaborating upon the most important facts needed to know that determine the context this research is embedded in. This includes an overview of all legally mandated responsibilities for all actors.
Chapters 5, 6 and 7 are the empirical chapters, which will work with the data gathered, answering first all sub questions of this research and thereby ultimately the research question itself. Chapter 5 will focus on the legal responsibilities of actors, chapter 6 on the practiced responsibilities. Chapter 7 then will discuss the power relations between actors and their effects on shaping them.
Chapter 8 concludes this thesis with a discussion that starts with a methodological reflection, to then give a final overview of the key variables influencing the development of the sanitation sector. The chapter finishes with recommendations for policy and practice.
2 Theoretical framework
2.1 Introduction
This chapter will explore the theoretical framework building the basis for analysis of power relations between actors. This theoretical framework allows us to systematically analyse the power relations that define why actors involved in the development of sanitation sector of Lilongwe either fulfil or not fulfil their legally mandated responsibilities, in other words if they practice them. I will first go through a conceptualization of power in general and then take a brief look at power in relation sanitation provision. Secondly, I will thoroughly introduce the power cube by Gaventa, and finalize this chapter with thirdly explaining the interrelationship of spaces, places and forms of power.
2.2 Conceptualizing power
There is a broad variety of conceptions about power in the academic world, such as seeing power as knowledge which implies that power is constituted through accepted forms of knowledge, or also that discourse transmits and produces power, through reinforcing it, undermining and exposing power (Cohen & Foucault, 1978; Foucault, 2003). For Foucault, power is present in all social relationships with its effects framing the boundaries for action, with different actors in policy processes finding themselves framed by the policy as discourse context (Brock, Cornwall, & Gaventa, 2001, p. 5), thereby standing in direction interrelation with the discourse they create through policy-making and the power implications for individual actors this carries. Gaventa refers to power in multiple ways and is thereby not limiting himself in the ways power can be understood. Gaventa elaborates that power can be understood as: power ‘over’ someone by affecting the actions and thought of the powerless; as power ‘to’ as capacity to act by exercising agency; as power ‘within’ referring to gaining the sense of self-identity, confidence and awareness as pre-condition for action; and lastly as power ‘with’, referring to the “synergy which can emerge through partnerships and collaboration with others (Gaventa, 2006b, p. 9). Governance on the other hand refers generally speaking to “the processes of decision-making by which society defines and handles its problems” (Iribarnegaray & Seghezzo, 2012, p. 2923). For government purposes, coercive, utilitarian and normative types of power may become relevant at some point or in specific places (Iribarnegaray & Seghezzo, 2012, p. 2924). The relation between governance and power is complex, and power “is just one of the attributes necessary to identify
relevant actors in multi-stakeholder processes, next to e.g. legitimacy or urgency among other criteria” (Iribarnegaray & Seghezzo, 2012, p. 2925). The definition of power that will be used to create a common understanding of power in a time-limited interview situation between the researcher and the interviewees is coming from Max Weber: “Power is the probability that one actor within a social relationship will be in a position to carry out his own will even against resistance, regardless of the basis on which this probability rests”(Weber, 1978, p. 53). The element of being able to carry out ones own will even against resistance provide examples of what Gaventa refers to power ‘over’, power ‘to’, power ‘within’ and power ‘with’ as explained earlier in this section. Weber’s definition thereby includes all ways in which power can be understood according to Gaventa. Its usability for a research setting of countries in the Global South has been proven through Schiffer using the very same definition in her research where she first used the Power Mapping Tool.
2.3 Power and access to sanitation in a development context
In the context of provision of sanitation, the capacity of urban poor groups to influence water and sanitation policies and providers depends on the responsiveness of government and providers themselves (David Satterthwaite & McGranahan, 2006). This manifests a clear relationship between users and providers, with power being in the hands of providers and users usually being dependent upon the providers. In an ideal world it can be argued that democratization and decentralization “ought to be particularly effective means of getting governments to be more responsive to water and sanitation demands”(David Satterthwaite & McGranahan, 2006, p. 20). The relevance of power relations between actors is obvious, as democratization and decentralization come hand in hand with increases in accountability, actor agency and the opening of new governance spaces that create a sharing of power amongst more actors, including the public. Through decentralization power is given to locals, and democratization allows for the inclusion of all members of society, including the urban poor. Yet, in many cities of the Global South, sewerage access remains highly stratified along class divisions, thereby “highlighting discrepancies in economic and political power”(Morales, 2016, p. 63). This stratified access to sanitation thus carries yet to be further discovered impacts on social cohesion between members of society along class divisions, and hinders economic and thereby eventually societal development and inclusion into society within areas having low access to sanitation. It can be argued that access to centralized sewerage systems has been used to “demonstrate a societal sense
of power over, and separation from nature”(Morales, 2016, p. 67). Furthermore, “sewerage serves as a tangible sign of state recognition”(Morales, 2016, p. 68) with its absence making inequalities in political power and citizenship highly visible, with Swyngedouw calling differential access to infrastructure a mechanism of domination and subordination, due to its ability to make power relationships visible (Swyngedouw, 2004).
2.4 The Power Cube by Gaventa
It becomes clear that power is extremely important and also difficult to research, even more so in a short period of time. It is for this reason that the Gaventa Tool, already academically established, has been used for this research as a tool for researching power relations. The power cube by Gaventa has been initially designed for analysing spaces for civic engagement within political processes, opening up possibilities for understanding and analysing current and past situations, in order to improve them through taking participatory action. As Gaventa notes the power cube shall be seen as an analytical tool, that is able to “assess the possibilities of transformative action in various political spaces”(Gaventa, 2006a). It furthermore allows focusing on the desirable variable of inclusion of all actors within political and generally decision-making processes, by being a conceptual framework that allows a simple distinction between different spaces, levels and forms of power. This conceptual framework has been used during the research as a tool to categorize and analyse the power relations between actors in Lilongwe, and has remained open, as argued for by Gaventa, for reconfiguration and adaption to the local context by the researcher.
2.4.1 Spaces of power
Spaces “are seen as opportunities, moments and channels where citizens can act to potentially affect policies, discourses, decisions and relationships that affect their lives and interests” (Gaventa, 2006a, p. 26). It is important to take into account, that spaces for participation are not neutral, but are themselves shaped by power relations which both surround and enter them (Cornwall, 2002; Gaventa, 2006b). Generally, power relations can help to shape the boundaries of participatory spaces. Within this analytical framework, a distinction can be made between three different kinds of spaces (Gaventa, 2006a, pp. 26–27):
In closed spaces, decisions are made by a set of actors behind closed doors, without any pretence of broadening the boundaries for inclusion. Some also see closed spaces as ‘provided’ spaces, where the elites makes decisions and provides services to the people.
Invited spaces are those into which people are invited to participate by various kinds of authorities, be they government, supranational agencies or non-governmental organisations –with the increase of participatory governance these spaces are seen at every level, from local government, to national policy and even in global policy forum. These spaces can be and usually are regularized.
Claimed spaces are spaces which are claimed by less powerful actors from or against the power holders, or created more autonomously by them. The spaces range from being created by social movements to community associations, going as far as the natural places where people gather to debate. During works with civil society actors in Colombia, researches identified an additional continuum of spaces that were either formal by invitation, formal by right, created by non-state institutions or civil society organizations. The concept of power spaces is yet to be further explored, and can be adapted to changing realities of power relations. A critical point here is, that those who create a space “are more likely to have power within it” (Gaventa, 2006a, p. 27). Another important point to be made is that “the transformative potential of spaces for participatory governance must always be assessed in relationship to the other spaces which surround them” (Gaventa, 2006a, p. 27). It is of crucial importance to recognize that these spaces always have to be seen in the interrelation of actors and players that created, shaped and designed them.
2.4.2 Forms of power
Gaventa identified three different forms of power, namely visible, hidden and invisible power. Regarding the different forms of power, Gaventa notes “we must also examine the dynamics of power that shape the inclusiveness of participation within each”(Gaventa, 2006a, pp. 28–29). After all, historically many pluralist studies of power have mainly examined power in its visible manifestations. However, according to Gaventa (2006), this does not reflect the whole spectrum of forms that power can take. According to him, the following forms of power exist:
Visible power is present when visible power is exercised in relatively open, observable decision-making processes. It can be defined as the visible and definable aspect of political power that is embedded in the formal rules, structures, authorities, etc. of decision-making.
Certain powerful people and institutions hold hidden power that reflects a capability of influencing and biasing e.g. who gets to the decision-making table, and what topics get on the agenda. Hidden power in short means setting the political agenda whilst remaining hidden.
Perhaps the hardest power to grasp is invisible power. It implies shaping meaning and what is acceptable, by shaping the psychological and ideological boundaries of participation. Invisible power is rooted in the socio-cultural normative system that is being adhered to, and influences how individuals think about their place in the world; it shapes peoples beliefs, the sense of self and their acceptance of a status quo. Changing this kind of power requires a focus on targeting the social and political culture as well as individual consciousness in order to transform the ways people see themselves within a larger system.
2.4.3 Levels of power
Lastly, Gaventa identifies three easily understandable levels of power. To keep it short, these levels of power can be identified to be the global, national or local level, whereby a more specific sub-categorization is not automatically implied, nor excluded. These three different levels of power are increasingly interrelated with each other, as in a globalized world global decisions are often strongly felt nationally or locally, and experiences made on the national or local level can often be used for informed policy decisions on a global level (Baker, 2013; Bavinck & Gupta, 2014; Gaventa, 2006b). This implies, that local forms and manifestations of power are increasingly interrelated with global actors and forces, and vice versa.
2.4.4 Interrelationship of spaces, places and forms of power
The spaces, places and forms of power shall not be viewed as independent, but rather highly interdependent from each other. As Gaventa notes the dynamics of power depends on all three dimensions of power, as well as successful strategies for change do. Furthermore, creating successful change is about aligning the different pieces of each dimension with each other. Just like with the Rubik´s cube, there are “literally billions of different positions that the blocks of the cube may have, illustrating the complexity and permutations in which power can take across space, place and form in any given context” (Gaventa, 2006a, p. 30). The image of the Rubik´s
cube illustrates the point being that any kind of “governance network can be defined as a horizontal articulation of inter-dependent but operationally autonomous actors who interact through negotiations that take place within a relatively institutionalized framework and facilitate self-regulated policymaking in the shadow of hierarchy”(Torfing, Peters, Pierre, & Sørensen, 2012, p. 16). As with the Rubik´s cube, which is solved only when all sides are aligned, change is most likely to happen when e.g. actors working on change are focussing their work onto all spaces of power, and not only in claimed spaces – thus when actors are aligned horizontally (Gaventa, 2006a). Similarly, aligning actors vertically from the global, to the national down to the local level carries more power than focussing on one level only. Lastly, the same applies for change strategies that try to horizontally align actors that use different forms of power, as it creates the most likely situation where power relations between actors are shifting, and new possibilities and developments in favour of change can occur (Gaventa, 2006a). It remains important to note, that the power cube approach does not give a recipe on how do induce change, but rather an analytical tool allowing actors to reflect where and how to do so, identifying partners for change and which elements of power dynamics to focus on.
2.5 Summary
This chapter has introduced the theoretical framework that is providing the theoretical background that this research is built upon. It has done so through conceptualizing power first in general, and then with a specific focus on power and access to sanitation in a development context. Lastly I explained the power cube approach by Gaventa, including the interrelationship of spaces, forms and levels of power.
3 Methodology
3.1 Introduction
This chapter discusses the methodologies being applied in this field research and analysis. I will start with introducing the research questions and sub questions of this research, followed by a discussion of epistemic assumptions and implications for the research design. Then, the research design itself is examined, as well as the unit of analysis and unit of observation being discussed. The conceptual framework follows, along with the operationalization table that is informed and built upon the conceptual framework. The sampling strategy is then elaborated upon, and within the methods of data collection section the power-mapping tool as developed by Schiffer is introduced in depth. Ethical considerations of this research are reviewed, before limitations for the research are highlighted. As a last step, a discussion is undertaken on how to best incorporate results from the power-mapping tool within the power cube concept from a methodological perspective.
3.2 Research question
Within this research question, several key concepts can be identified that require a clear-cut definition that will be used throughout this thesis in order to avoid any misunderstandings.
The Government of Malawi has defined sanitation as: “The provision or availability of structures that facilitate hygiene. The structures include, for example, water taps, boreholes, toilets bath shelter, hand-wash facilities, washing basins, slabs, soak pits and solid waste management structures (i.e. refuse pits, incinerators, compost heaps, sharps pits etc.). In other words, sanitation means to have the facilities (‘Zipangizo za ukhondo’)”(Ministry of Irrigation and Water Development, 2010). The sanitation sector therefore is the political, economic and social/cultural sector that is concerned with creating or improving access to sanitation to ever
How are legally mandated responsibilities and practiced responsibilities taken up
by actors for the development of the sanitation sector in Lilongwe, Malawi
differing from each other, and how are those responsibilities affected by power
more inhabitants of Lilongwe City. Within the framework of this research, the focal point of interest has been on sanitation relating to human fluid waste, and thereby also on infrastructures associated with providing hygienic environments for disposal of such fluid waste (e.g. sewerage systems or public toilets).
An actor in this research is defined as in the framework of policy analysis for sustainable development, which states that actors “contribute to a policy problem, are needed for solving the problem, or are affected by the problem-solving activities”(Runhaar, Dieperink, & Driessen, 2006, p. 41). While actors within this framework turn into stakeholders through fulfilling these requirements, I chose to use the wording actor consistently throughout this thesis, for sake of clarity and simplicity. Thus, actors are herewith and for this thesis defined as participants in an action or process contributing to the development of the sanitation sector in Lilongwe City.
Furthermore, a difference is being made in the research questions regarding legally mandated, and practiced responsibilities. They shall be defined as follows: Legally mandated responsibilities are all responsibilities and tasks that have been assigned to a specific person, or an organizational role within an organisation or institution (e.g. Director General of Lilongwe Water Board), on the basis of a legally binding document (national to local policies, contracts, laws etc.). Practiced responsibilities are responsibilities that are actually being carried out by actors on a practical level, regardless of the assignment of a legal responsibility. Practiced responsibilities can overlap fully or partially with the legally mandated responsibilities, but can also differ and be taken up by actors that have not been legally assigned to carry out that responsibility.
3.2.1 Sub-questions
In addition to the main research question, several additional sub-research questions have been deployed.
1.Who are the actors involved in the sanitation arena in Lilongwe, Malawi? 2. What are the legally mandated responsibilities of these actors?
3. What are the practiced responsibilities of the actors? What is actually happening on the ground?
4. What do the answers to the earlier sub-questions reveal about the interactions and power relations in place?
5. How are the power relations in place affecting responsibilities of actors?
3.3 Epistemic assumptions and implications for research design
This research is embedded in a critical realist approach, which holds that there is a reality that exists independent from the human conception. Applicable to both natural and social sciences, critical realism carries the belief that culture and society, which includes understandings of power, are created by human activities. Regarding social sciences, critical realism holds that the researchers always needs to be critical of its objects of study (Sayer, 2006). In contrast to natural laws, rules of culture and society are not generally applicable but are highly dependent on the geographic context and time. This implies, that critical realism is not able to predict, and that this theory can be used for explanatory benefits only. Critical theory requires that researchers go beyond the observable, in order to investigate the mechanisms behind any event. In the light of Gaventa´s power cube, and especially the forms of invisible and hidden power, critical realism is the only epistemology that can reasonably be adopted for this research. The level of analysis in critical realism can be the individual, the group or an organization.
3.4 Research design
This field research is designed as a mixed methods study, and uses both qualitative and quantitative methods in order to increase the quality of the results (Johnson & Onwuegbuzie, 2004). Naturally, using multiple methods reduces the risk of false results due to weaknesses of one approach. As pointed out by Bryman, the integration of data from different methods, and not separate handling of data safeguards that the benefits of compensation for weakness of one method are ensured (Alan Bryman, 2008). Additionally, using a mixed methods approach allows the researcher to triangulate the collected data, thereby finding out whether findings stemming from different methods are convergent (Johnson & Onwuegbuzie, 2004). Lastly, similarity of results from different methods increase the validity of the research (Guion, Diehl, & Mcdonald, 2011). Throughout this research, preliminary quantitative data gathered has been triangulated with results gathered through qualitative methods. Furthermore, most important qualitative data gathered has been crosschecked with other respondents in different settings.
3.5 Unit of analysis and unit of observation
The unit of analysis for this thesis proposal are the power relations between actors responsible for the development of the sanitation sector in Lilongwe. A hardly directly measurable unit of
analysis, these power relations are represented by perceptions of power that actors within the sanitation sector have from each other. The unit of observation of this research are the actors, which are taking up the legally mandated and practiced responsibilities that drive the development of the sanitation sector in Lilongwe.
3.6 Conceptual Framework
This conceptual framework has been guiding the data gathering process and shaped its analysis. It has to be understood and read in the following way. Starting in the left top, this conceptual framework argues that actors of the sanitation sector in Lilongwe are all embedded in certain power relations. These power relations are conceptualized and analysed on the basis of the Power Cube by Gaventa as introduced earlier, which has forms, spaces and levels of power as key distinctive elements for analysis. This conceptual scheme furthermore argues that the power relations that all the actors of the sanitation sector in Lilongwe, Malawi are embedded in are
shaping the legally mandated and practiced responsibilities the actors are supposed or possibly also not supposed to take up. It is important to note, that the sphere of legally mandated responsibilities may fully, partially or not at all overlap with the sphere of practiced responsibilities taken up by actors. The two circles with the arrow on the right side illustrate this potential overlap. Actors for the development of sanitation in Lilongwe City can carry both the legally mandated as well as practiced responsibilities to varying degrees, depending on the actor looked at. The fact that the actors carrying the responsibilities are embedded in power relations signify a feedback loop, as the responsibilities taken up by actors also eventually shape their position of power in relation to other actors and vice versa.
This conceptual framework provides clarity of the hard to grasp concept of power, and how it interrelated and interacts with various social actors. I observed that cultural differences manifested themselves in the way that interviewees living in an entirely different cultural environment can perceive power. Given the differences of how social relationships are lived in Europe in comparisons to Sub-Saharan African countries like Malawi where the research took place, a coherent definition of power proved to be a red thread that people could stick to when thinking about the ways in which power is exercised and shaping practiced and legally mandated responsibilities. This coherent definition of power, as was given throughout the interviews by me to the interviewees, is of crucial importance for establishing a baseline of collective understanding that is required for coherence of data. To remind ourselves, power to be understood within the framework of this thesis, is defined by Max Weber: it is the “Probability that one actor within a social relationship will be in a position to carry out his own will even against resistance, regardless of the basis on which this probability rests”(Weber, 1978).
3.7 Operationalization table
Concepts Dimensions Variables Example Indicators
Data Sources
Power
Forms
Visible
Power exercised in
relatively open, observable decision making processes
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/PMG/Field
observations
Hidden
Controlling topics that form the agenda / Influencing who gets to join the decision making table
Interviews (Focus Group, Face-to-face)/ PMG/ Field observations
Invisible
Influencing how
individuals think, what they belief, there sense of self and acceptance of status quo
Interviews (Focus Group, Face-to-face)/ PMG / Field observations
Spaces
Closed Decisions made behind closed doors YES/ NO
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/ PMG
Invited Invitation to participate given YES/ NO
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/ PMG
Claimed/Cre ated
Space created by less powerful actors YES/NO
Interviews (Focus Group, Face-to-face)/ PMG / Field observations
Level
Local
Power level of actor stemming mainly from local level YES/NO
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/ PMG/Field
observations
National
Power level of actor stemming mainly from national level YES/NO
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/ PMG /Field
observations
Global
Power level of actor stemming mainly from global level YES/NO
Legal documents, formal rules and structures of decision making
processes / Interviews (Focus Group, Face-to-face)/ PMG /Field
observations
Based on the research question and the conceptual scheme, three interdependent concepts can be identified. These are firstly power and the relations it has within and between the dimensions it
can take, legally mandated and practiced responsibilities and lastly actors of the sanitation sector in Lilongwe, Malawi. The operationalization table consists of one major concept, namely power, being the only concept that requires a thorough operationalization. The concepts of legally mandated and practiced responsibilities, as well as sanitation, the sanitation sector itself and actors within the sanitation sector have already been defined in chapter 3.2.
3.8 Sampling strategy
In order to align the sampling strategy with the research question of this research, nonprobability-sampling methods are applied in this study. In furtherance of receiving a broad variety of respondents from different actor organizations and individual actors, the methods of purposive sampling and snowball sampling have been applied. Purposive sampling implies, that the researchers specifically targeted and approached actors and interviewees that are of interest for the study, with all respondents being interviewed (Alan Bryman, 2008). Furthermore, the strategy of snowball sampling has been followed, which implies that existing study subjects helped connect to more subjects into the sample (Lynch, 2008). Last but not least, accidental sampling has been followed as a strategy wherever applicable, which means that the sample is being drawn whenever readily available and convenient (Alan Bryman, 2008; Lynch, 2008).
3.9 Methods of data collection
Three main methods of data collection have been deployed for this research. Firstly, a literature review has been undertaken before and after going to the field. Secondly the PMG has been used for gathering quantitative data about perceptions of power. Thirdly, semi-structured interviews have been carried out with actors in the field, which implies that on basis of qualitative research standards an interview guide has been developed for asking the right questions (A Bryman, 2012). These semi-structured interviews have been, where applicable, carried out after the PMG has been applied already, to support but also question quantitative findings with qualitative responses as well as to ask questions within the framework of Gaventa’s power cube.
3.9.1 Literature review
This research required a literature review with regards to how power can be conceptualized, and especially how it can be measured. Outcomes of this literature review included findings that power is a highly complex concept, wherefore a single definition of power is required for a
collective understanding of the issue that roots findings in a common denominator. The second element requiring a literature review was of legal form, in the sense that legal documents needed to be reviewed, in order to shed light onto the legally mandated responsibilities on the ground. Unfortunately, not all necessary documents were available online, but some additional documents were gathered whilst talking to government officials who were so kind to send them on to me. 3.9.2 Power mapping game (PMG)
In order to study the relations of power between actors, the PMG as developed by Schiffer in 2007 has been deployed. It allows contributing “structured qualitative and quantitative data concerning the perception of power of the different actors involved”(Schiffer, 2007, p. 3). The PMG by Schiffer is specifically useful and applicable for this research, as it is an empirically based analysis of power on the ground in contrast to a purely theory-driven general analysis. It thereby is able to analyse power relations from a perspective that is reality for actors involved, as it measures perceptions of power as perceived by actors themselves. The tool has been developed for the direct and “hands-on analysis of power structures with the participation of actors involved in the governance field”(Schiffer, 2007, p. 3), and is adapted to specific requirements for research countries of the global south, such as for example high illiteracy rates, high diversity of research partners and interviewees in terms of culture and language and restricted technical infrastructures. The power mapping tool and the way it works can be best explained in six steps that will be explained throughout the following section. The equipment necessary for application of the PMG, as well as the power matrix for noting down results can be found in Appendix A. All information presented within this section is heavily based on the methodology as developed by Schiffer (2007), with minor adoptions made for this research. 3.9.2.1 PMG: Six steps of data collection
Assembling actor list and choosing figures
In the beginning of the PMG, the interviewee is asked to “list all individual or organizational actors that s/he sees as involved in a specific governance field, activity or organization” (Schiffer, 2007, pp. 9–10). For this research, this was the field of actors responsible for the development of the sanitation sector in Lilongwe. This list is not limited to formal decision-makers, but also includes other groups such as users, donors, civil society organisations, etc. All these actors are noted down in the power matrix list as seen in Appendix B. For all actors the interviewee had to
choose one game figure with their distinguishing characteristics (red, blue, yellow, black) also noted down in the second column of the power matrix.
Drawing governance map and arranging actors on the map
In a second step, a governance map is drawn out (which is not a geographical map) that represents a sketch of all the actor groups involved. Once the map is drawn and agreed upon, all actor figures are arranged on it following the assembled actor list from the first step. It was expected and later on confirmed that the maps would look different per interviewee, based on differing perceptions on which actor is the most powerful for the sanitation sector in Lilongwe. Wherever an actor is member of more than one actor group, an intersection occurs between two power fields.
Adding of range-of-action-cards/coins
Each actor is influencing his field through various activities. S/he can be either an observer, give advice, make decisions, fund, implement or legally enforce. The initial list has been expanded during the first interviews with the activities of implementation and legal enforcement as they surfaced as responses. ‘Traditionally’, each actor is characterized by the interviewees through putting down one or more of the range-of-action cards. Within this research, the addition of range of actions has been undertaken through ticking the respective boxes on the power matrix sheet only, and not through addition of cards on the sheet itself for reasons of simplicity and oversight. Additional comments made during this phase are noted down in the last column of the matrix, as it provides essential additional qualitative data to be worked with.
Defining power and arranging power towers
In order to receive coherent results, it is essential that the interviewer and the interviewee share one understanding of power. Thus, at the beginning of each interview the definition of power by Weber (1922) as stated already in chapter 2.2 has been given to the interviewees, stressing sources of power can be diverse. As Schiffer notes, “power is described as the ability to reach one’s goals in a social situation and the following question that has been asked is: “How much power does this actor have in this specific field/activity/organization” (and not in a more general sense)”(Schiffer, 2007, p. 31). Following a shared understanding of power, the interviewees have
been asked to rate how much power an actor has in the field of development of the sanitation sector in Lilongwe. Each actor thus then has a tower of checkers pieces with the following rules applying:
“The more powerful an actor, the higher the tower (more checker pieces stacked upon each other)
• The towers can be as high as interviewees want • Two actors can get the same height of power tower
• The power of actors in one actor group does not add up.”(Schiffer, 2007, p. 32)
After the identification, grouping and characterization based on range of action and relative power status of all actors, the researcher verbalizes the setup and encourages the interviewee to adjust the picture wherever necessary. When the interviewee is satisfied with the set-up, “the height of the power-towers is noted down in the fourth column of the matrix”(Schiffer, 2007, p. 31).
Discussing: Why and how
Until now, the researcher has mainly gathered quantitative data, with answers received to close ended questions that were visualized in a board-game manner. This setup will now be used to start a governance situation discussion, which will depend on the map created and the interviewee/s involved. Example questions that can be asked here are:
• “ “I see you have put this one on the highest tower. Why? Where does his/her power come from? How would someone from outside like me see that? “
• “You say that these two have the same power. What happens if they disagree? Is their power based on the same grounds? Does it have the same range?”
• “I have heard there is a conflict about xy between these three organizations. Could you explain to me, what is happening there, how do they use their power in the conflict?” • “You say this actor has no power at all. Why is that so? What could he do to get more
Lastly, the PMG also allows discussing the sensitive issue of power relations in a more abstract way, such as: “People tell me this blue figure has a problem with this yellow one. Tell me, what is it all about?”
3.9.3 Semi-structured interviews with actors involved in sanitation development
The goal of this research is to understand how power relations are shaping the assumption of legally mandated and practiced responsibilities of actors for sanitation development in Lilongwe. Hence, in depth qualitative interviews were required, in order to be able to explore the power relations through the lens of the concept of the power cube as developed by Gaventa. In almost all cases, the in-depth semi structured interviews were carried out in a face-to-face manner, with only three cases of interviews where more than one respondent was present. Gathering interviewees happened through snowball sampling, with interviewees giving additional contacts for further research (Alan Bryman, 2008). All interviews were recorded, and notes were taken where recording was not permitted or not feasible due to the sensitivity of the topic.
3.10 Triangulation
To ensure validity of results, I have used the method of respondent validation to triangulate quantitative and qualitative data of respondents with each other. I have done so by presenting the findings to other actors, specifically asking about claims made by one respondent. Given the differences of perceptions of power, triangulation always took place as last part of the interviews, in order to ensure a collectively shared understanding of how power is to be understood within the framework of this research.
3.11 Mode and methods of data analysis
For the analysis of the data, several modes and methods of data analysis have been undertaken. The first source of data gathered is stemming from the PMG. The information recorded has been transformed into sets of actor data, with one set of actor data consisting of what one interview partner says about one actor (e.g. information gathered in one line of the power matrix) (Schiffer, 2007, p. 32). Furthermore, for allowing comparison between different interviews, each “absolute figure representing the power status of one actor in one interview is turned into a relative power value: For each interview the highest tower is set equivalent to one and the ground level is set equivalent to zero”. This means that “each actor is assigned a relative power value between zero and one (by dividing the power value of this actor by the value of the highest tower in this
interview)” (Schiffer, 2007, p. 33). Each actor received an overall average aggregate value that is based on all power values given to this person by other actors. Furthermore, the main type of activity as described by the respondents (observing, giving advice, funding, implementing, legal enforcement, making decisions) has been included in this data, which has been analysed and visualised using Microsoft Excel. Additionally, group belongings have been attached to each actor, which allows zooming out with regards to larger groupings of actors, and the ways they act as group. From all the PMG’s that have been carried out, a master map of all actors for the development of the sanitation sector in Lilongwe has been created.
The second kind of data stems from the semi- structured interviews taking place around the PMG. Here, the most important information has been transcribed, and fed into Atlas TI using open coding and later on selective coding in order to find out the most important and reoccurring research themes of interest.
Thirdly, field observations notes and notes from informal conversations have been transcribed, and used to create a more holistic picture of the things happening on the ground. All three sources of data are combined and analysed in context with each other, in order to interrelate the interview and observation findings with the power game findings. Preliminary findings from in the field have been triangulated through respondent validation in the last two weeks of the research in the field. A specific focus has been put onto the ten actors that have been named most often during the PMG. Furthermore, in order to evaluate correctly whether or not actors are assuming their legally mandated responsibilities, an evaluation system of the amount of actually practiced responsibilities has been developed. This evaluation is based on the comparison of legally mandated responsibilities with data gathered in the PMG, semi-structured interviews and field notes taken. Values are assigned to each actors actual practice of legally mandated responsibilities on an ordinal scale that is reaching from “Not at all” to “Fully” with three evenly distanced steps in between. Finally, all data gathered is analysed using Gaventa’s power cube, and a special focus is taken on the forms, spaces and level of power involved in the development of the sanitation sector.
3.12 Ethical considerations
Given the sensitivity of the issue at hand, namely power relations, utmost importance has been attached by the researcher to the issues of confidentiality of data and information gathered. In order to avoid limiting respondents in their answers, the researcher ensured that answers will
remain absolutely anonymous, and cannot be traced back to persons that have said it. All respondents of this research have been handed out an informed consent form, that informed the respondents of the ways in which their data may be used, how it will be analysed and who it will be made available to. Furthermore, it informed the interviewee that all answers are voluntary and will be made anonymous. Lastly, no data has been gathered and published without a signed informed consent form. For some exceptional instances that remained as limited as possible, e.g. field observation notes or informal interviews where a signed informed consent form was not always readily available or simply not feasible within the conversation flow, the researcher made sure to inform the participants orally about the goals and purpose of this research. Also, during focus group interviews privacy of the responses could not be ensured, and all participants have been informed that they should not disclose the information that has been talked about during the group meetings to others.
3.13 Limitations
This research carries some limitations that the researcher is explicitly aware of. Firstly, results are not scientifically generalizable, as they are highly dependent on and shaped by the local context of the sanitation sector in Lilongwe, Malawi. Secondly, even though all efforts have been undertaken to minimize skewed results, I was only able to measure the perception of power of actors regarding their own and others power, including their relation to another, but not the “real” situation itself. After all, power is an intangible element of social life and governance that cannot be satisfactorily measured directly. Secondly, another critical point to be made lies in the fact that the conceptual framework argues responsibilities are defined by power relations and vice versa. The framework is limited in the sense that its focus lies on power itself, crossing out possible side-variables that may define or even shape power relations. Alternative variables could for example include actors influencing power relations from the outside, such as a large corporations investing into Malawi for economic reasons and thereby bringing in funding. Hence, while correlation can be proven between power relations and responsibilities taken up by actor, no definite causation chain can be established with certainty. Possible mitigations for these limitations cannot be reasonably offered. However, I believe the strength of this research lays its specific setup with results important and usable for the sector this research has been embedded in. While results may not be generalizable, methodological approaches such as combining the PMG with the power cube framework can be exported to other settings of research.
3.14 Combining the PMG with the power cube
One of the key ideas of this research was to combine the PMG with the power cube approach for analysing power. While the PMG is extremely useful for gathering data in a development setting, the power cube approach proves to be useful to categorize the levels of power, and to understand and conceptualize the concepts of forms of power, in addition to spaces of power where development takes place. The Power Cube furthermore provides a valuable tool for analysis for spaces of change, in order to evaluate where efforts for development will be most fruitful and beneficial for all. Within this research, both approaches are combined as results from the PMG are taken into account for the analysis of power relations between actors, in addition to looking at the forms, spaces and levels of power with and within which these power relations are taking place. This combination of approaches is carried out in an analysis of power relations and their effects on actors for taking up responsibilities in Chapter 7.
3.15 Summary
The purpose of this chapter is to inform about the methodology that has been guiding this research. This chapter presented the main research question and its sub questions, as well as the research design and conceptual framework that guided all research activities undertaken in the field. Methodological topics such as sampling strategy and methods of data collection have furthermore been discussed. Most importantly, this chapter explained the PMG in depth, and the strategies for methods used for data analysis. Lastly, the combination of the PMG with the power cube has been shortly introduced.
4 Context Setting
4.1 Introduction
Within this section, the necessary information is given to the readers about contextual factors relevant for this research. This foremost includes an introduction to all actors involved in the sanitation sector in Lilongwe City and their current strategies for improvement and development of the sector. An in depth exploration of their legally mandated and practiced responsibilities is given in chapters 5 and 6, and the effect of power relations upon these responsibilities is analysed in chapter 7.
4.2 National context
Malawi has a population of 17,3 million with the number of inhabitants growing by 3,3 % each year (Gieler, 2016; World Bank, 2015). Its capital, Lilongwe City has a population of 867,000 inhabitants and thereby is the most populated city in the country. Within Malawi only 16% of the population live in cities and 84% live in rural areas (Gieler, 2016; ISF‐UTS, 2011). A DfID official supports these statistics: “a lot of people live in rural areas. That’s the strength. We reach a higher amount of people”. This setup helps to explain why most development projects and funds are being channelled into rural areas, as the necessity and impact are evaluated as stronger by development partners. While poverty is focussed in rural areas the greatest disparity between rich and poor is in urban centres (WaterAid Malawi, 2010), furthering societal divide along class lines and creating more visible inequalities with regards to access to sanitation services. The GoM openly acknowledges that service provision of both water and sanitation are inequitable, particularly sanitation with coverage ranging between 12% and 90% in different districts (ISF‐ UTS, 2011). It has to be stressed once more though, that these statistics carry only limited reliability, based on the varying definitions actors use with regards to sanitation. The average income per person is 142 USD a year (2014), which makes Malawi one of the poorest countries in the world. Within the Human Development Index (HDI) Malawi is number 174 from 187 countries, and in 2005 52% of the population was judged to be “poor” i.e. below the poverty line of 1,25USd; 22% of the population were judged “ultra poor” – thus so poor that they cannot even afford to meet the recommended minimum daily food requirements (Gieler, 2016; UNDP, 2015; WaterAid Malawi, 2010). Naturally, sanitation is not a main concern when there is not even enough money for food – increasing pressure onto the GoM to provide its citizens with the
