Researching with care: ethical dilemmas in co-designing focus group discussions

Abstract

This paper reflects on the ethics of research and knowledge co-production aimed at addressing urban inequality. It draws on work within the Knowledge in Action for Urban Equality (KNOW) programme, which aimed to co-produce knowledge to activate transformation. We employ a lens of feminist care ethics to examine ethical challenges in research partnerships, which derive from interrelated layers of power asymmetry and inequality. We focus on ethical dilemmas that emerged during the planning stage of research work led by the NGO Centre for Community Initiatives (CCI), Tanzania, in collaboration with University College London’s Institute for Global Prosperity (IGP). We argue that contextualizing the value of knowledge co-production in generating transformation in the long term reveals a necessity for simultaneously addressing the immediate needs of intersectionally marginalized research participants. We suggest that ethical awareness of both long- and short-term modes of “caring for” could better support initiatives for addressing urban inequalities in context.

I. Introduction

This paper reflects on the ethics of research practice, focusing on urban inequality, intersectionality and care. It draws on work conducted within the Knowledge in Action for Urban Equality (KNOW) programme (2018–2022),⁽¹⁾ the aim of which was to co-produce knowledge for activating transformation and removing structural barriers that lie at the root of inequality. In this paper we look at ethical challenges in research partnerships, which emerge from interrelated layers of power asymmetry and inequality operating at various scales: from the global scale of international research partnerships, to the institutional scale of academic and non-academic organizations, partners and researchers, and then to the local scale, where researchers and their interlocutors, as well as research participants, interact. The ethical dilemmas we focus on emerged during the planning stage of research work led by the Centre for Community Initiatives (CCI), an NGO in Tanzania, in collaboration with University College London’s Institute for Global Prosperity (IGP). Their research collaboration aimed to understand and measure situated perceptions of prosperity in informal settlements in Dar es Salaam, and to use the findings for action towards transforming urban policies.

This paper examines some of the ethical debates that came up during this research around the expected impact of focus group discussions on the participants, the researchers and the data collected. These debates highlighted the ways in which issues around gender equality intersected with differences of class and income-generating status among the participants. We examine how these issues affected the structuring of the focus group discussions as a participatory research method for knowledge co-production. Understanding the internal dynamics, which are shaped by the power relations and different priorities of multiple stakeholders, is crucial for co-producing knowledge. It raises questions of benefit, and highlights the difficulty of balancing the overall project aim of achieving long-term policy changes on the one hand, with the immediate needs of vulnerable groups on the other. While the contextualizing of knowledge co-production aimed at increasing urban equality is a powerful tool for achieving long-term transformation, we argue that it requires a simultaneous addressing, in the short term, of the differing needs of the co-producing stakeholders, particularly the urgent needs of intersectionally marginalized participants.

II. Co-Producing Knowledge For The Prosperity Index In Dar Es Salaam

The paper is based on empirical work conducted in a series of workshops at CCI in Dar es Salaam, Tanzania, in November 2019. CCI was established in Dar es Salaam in 2004, with the aim of providing technical and financial assistance to communities living in informal settlements, and to the Tanzania Urban Poor Federation (TUPF). CCI works to build the capacities of communities to initiate sustainable development that will improve their livelihoods, to facilitate the development of organizational capacity at the local level, and to promote pro-poor policy and practice.⁽²⁾

CCI has long working relationships with communities, as well as with various levels of governance, academia and other authorities in Dar es Salaam, and is experienced in community organizing, facilitation and participatory research. In the context of the KNOW project, CCI has centred its research on several thematic topics, including wastewater, solid waste management, urban risk and prosperity analysis. In the research described here, the collaboration of NGO and academia facilitated the creation of what Henry Myrttinen and Subhiya Mastonshoeva refer to as a “mechanism of epistemic connections”,⁽³⁾ through the co-development of a research plan by the local team of researchers. Their research aimed to develop a local Prosperity Index as a tool for communities to influence urban policymaking, and to generate pathways to increasing urban equality.⁽⁴⁾ This methodology was developed by UCL’s Institute for Global Prosperity (IGP), one of the KNOW project partners, and introduced to the team in Dar es Salaam by Dr Saffron Woodcraft from IGP⁽⁵⁾. The IGP defines the Prosperity Index as “a mixed-methods community co-production process, led by residents working in partnership with academic researchers and NGOs, to address the lack of context-specific policy-relevant knowledge about prosperity”.⁽⁶⁾ The training sessions discussed in this paper focused on the first stage of the research, in which the local research team planned a series of focus group discussions, aimed at defining a situated model of prosperity.⁽⁷⁾

III. Analyzing Co-Production Processes Using Intersectionality And Feminist Care Ethics

The ethical dilemmas discussed in this paper relate to wider debates around questions of addressing power relations within the co-production process, in order to avoid reproducing structures of exclusion and reinforcing intersectional marginalization. We question whether, and to what extent, the lived experiences of social groups facing multiple layers of marginality and exclusion can be recognized, acknowledged and incorporated into knowledge co-production processes. We ask how co-production processes can be harnessed to disrupt social categories that shape these power relations and affect subjectivity-making.⁽⁸⁾ We rely on intersectionality theory and feminist care ethics as critical lenses through which to understand co-production and its complications in dealing with multiple and intersecting layers of power and vulnerabilities.

Knowledge co-production is widely discussed in the literature as a method of working together with partners to jointly define research questions and generate new knowledge, projects or products.⁽⁹⁾ The process of co-production involves collaborating with different stakeholders so as to deliver an outcome or process that is grounded in the relevant social, cultural and political context, and that incorporates the various ways in which different stakeholders approach, understand and deal with research questions.⁽¹⁰⁾ Processes, practices and products of co-produced knowledge can integrate different perspectives, and also have the potential to be transformational, becoming agents of change that can affect different stakeholders.⁽¹¹⁾ In this way, knowledge co-production enables a form of social learning that can challenge existing assumptions that might otherwise prevent transformative change.⁽¹²⁾ The social effects include the gaining of new knowledge, but the process can also encourage network building, increase public involvement, lead to the development of a wider understanding of different perspectives, and enhance decision-making capacities.⁽¹³⁾ Common goals and purposes must be carefully defined, and the details of working together should be agreed upon at the outset of the research, to ensure that all partners gain something from the project.⁽¹⁴⁾

Since co-produced research is performed by multiple stakeholders, at the core of this method are the interactions and communication between partners. This requires attention to the centrality of ethical principles such as recognition and respect for the knowledge and value systems of the various stakeholders.⁽¹⁵⁾ Simin Davoudi discusses the incorporation of multiple forms of knowledge as a way to reconceptualize the notion of evidence and extend its validity beyond the realm of experts.⁽¹⁶⁾ Furthermore, as outlined by Vanesa Castán Broto and Susana Neves Alves, co-production must incorporate the different and varying identities that continuously shape the dynamics of social life, since “[c]o-production is one such means of negotiating both identities and the rights and obligations associated with those identities”.⁽¹⁷⁾

These ethical aspects are central to co-producing partnerships between academic and non-academic researchers and touch on questions of differences that inform ways of working, both epistemological and practical. Daniel Stevens et al. point out that multi-sectoral collaborations involving academics and NGOs require negotiating through these differences, and emphasize that “academics and NGO practitioners often hold different worldviews and have different levels of comfort with, and confidence in, established analytical categories”.⁽¹⁸⁾ Ronan McDermott et al. assert that “[p]articular ethical considerations arise in such projects due to the varying mandates, objectives and ways of working of the organisations involved”.⁽¹⁹⁾ Such differences can lead to tensions, for instance around issues of power relations between partners; allocation of time and resources; the need to balance different expectations and interests; and ways of working through disagreements, misunderstandings or conflicts. Problematic attitudes may surface, such as competitiveness, ethnocentrism or paternalism.⁽²⁰⁾ In such contexts, even the discussion of ethical issues could tend towards re-embedding colonial ideas concerning relationships, respect and responsibility.⁽²¹⁾ To avoid this, co-producing research partners need to identify and acknowledge intersections of disadvantage as they formulate their relational responsibilities.

The term “intersectionality” was first introduced by Kimberlé Crenshaw with reference to how inequalities that are connected to social categories such as gender, class, race, ethnicity and other social divisions can intersect, producing multiple and overlapping dimensions of domination, subordination and discrimination.⁽²²⁾ In particular, feminist scholars have used intersectionality as an approach in feminist theory for examining the causes of multiple sources of women’s oppression, related to class, ethnicity, race and sexuality as well as gender and sex.⁽²³⁾ Intersectionality theory also refers in more general terms to the experiences of social groups located at those intersections.⁽²⁴⁾ In the context of urban inequality, an awareness of intersectionality is crucial for analyzing social relations of exclusion. As noted by Castán Broto and Neves Alves, it provides “a means to claim multiple sources of oppression that affect vulnerable communities and that are not always represented in feminist analyses of urban inequality”.⁽²⁵⁾

We argue here that feminist care ethics is a useful lens for considering how to acknowledge, incorporate and address intersectional marginalization within co-production processes, by offering ways to think about the recognition of, response to and reception of care. Joan Tronto outlines four phases of care:

caring about – i.e. recognizing a need for care; caring for , i.e. taking responsibility to meet that need; care giving , i.e. the actual physical work of providing care; and, finally, care receiving , i.e. the evaluation of how well the care provided had met the caring need.⁽²⁶⁾

Care ethics emphasize the importance of dependencies and relationships between human beings. This suggests an alternative way of looking at the universal ethical principles represented by approaches such as deontological ethics, utilitarianism and consequentialism, all of which are influenced by modernist Western liberal human rights theory. Virginia Held points out that:

In the dominant moral theories of the ethics of justice, the values of equality, impartiality, fair distribution, and noninterference have priority; in practices of justice, individual rights are protected, impartial judgments are arrived at, punishments are deserved, and equal treatment is sought. In contrast, in the ethics of care, the values of trust, solidarity, mutual concern, and empathetic responsiveness have priority. ⁽²⁷⁾

Hilde Lindemann points out that since care work, both paid and unpaid, is overwhelmingly done by women, the notion of care ethics has been criticized on the grounds that it could reinforce gendered stereotypes. Further critique has pointed to tensions between the focus of care ethics on the particularity of interpersonal relationships, and the need to broaden its scope to challenge and offer alternative ways to address a wider sociopolitical context.⁽²⁸⁾ Maria Puig de la Bellacasa warns against uncritically embracing the notion of care ethics, because of its possible integration into hegemonic ethics as a tool of normative moralization. Instead, she argues: “the ‘ethics’ in an ethics of care cannot be about a realm of normative moral obligations but rather about thick, impure, involvement in a world where the question of how to care needs to be posed”.⁽²⁹⁾

“Unsettling” care is also the focus of Michelle Murphy’s warning that the transnational politics of care might potentially “work with and through the grain of hegemonic structures, rather than against them”.⁽³⁰⁾ Her critique of postcolonial legacies is central to the question of how different types of “care giving” and “care receiving” need to be conceived and implemented against a background of critical awareness of colonialism.

In our analysis we ground such dilemmas in a situated context, using an intersectional feminist lens that allows for a more nuanced understanding of care and its application within a knowledge co-production process. The team that participated in the Dar es Salaam workshops was conscious of how care is situated within contexts of power and control. Some of the focus groups they had initially suggested were intended to include a range of participants, some of them more vulnerable (including unemployed youth and people with disabilities), others less so (such as local leaders, entrepreneurs and landlords). Possible points of tension that the team anticipated were derived from the many differences among the collaborating stakeholders, including disciplinary, institutional, cultural and class differences, resulting in power imbalances. In the following sections, we comment on the social performance and the differing positionalities of partners within the research team, and explore their contrasting views regarding intersecting differences of gender, class, income-generating status and professional knowledge among the research participants.

IV. Methodology Layers

We conducted participatory observations of the research design process over a week of workshops in Dar es Salaam. Two of the sessions during this week were dedicated to discussions of research ethics, in which we initiated discussions of ethical problems and difficulties with the research team who also suggested ways to address them. Following a dilemma that emerged around the formation of focus groups, which we explore in detail in this paper, we also held short open-ended interviews with two members of the research team who expressed contrasting views on this dilemma.

Beyond these methods of gathering data, in this paper we reflect on additional interrelated layers of methodology that were employed in the research: knowledge co-production as the overarching methodology of the KNOW programme; focus group discussions as part of the Prosperity Index methodology developed by the IGP; and the interpretation and adaptation of these methodologies to the situated local context by the multi-stakeholder research team assembled by CCI in Dar es Salaam.

This paper investigates the ethical issues at stake in practising such a multi-layered methodological mode of co-production. We follow Lindemann’s understanding of feminist ethics, which she argues is not a branch of ethics, but rather a way of doing ethics.⁽³¹⁾ This approach complicates the distinction that is often drawn between theory, method and practice. The blurring of these boundaries is articulated by D Soyini Madison in her discussion of critical ethnography, where she observes that “[t]heory, when used as a mode of interpretation, is a method”.⁽³²⁾ We follow her suggestion to “think of ethnography as critical theory in action ”,⁽³³⁾ when we reflect on the practice of ethics.

V. Designing The Research

We begin by looking at the formation of the local research team, and the ways in which the team made decisions about how the participatory research would be conducted. CCI carefully assembled a diverse research team that brought together people from various social positions. The main considerations for selecting the participants were to include people who are affected and impacted by poverty on the ground, as well as to bring in government officials who are involved in planning, local government officials, and CCI research team members, who were the facilitators.

The formulation of the team, which took into account a gendered and intersectional approach, was therefore an integral part of the process of co-producing knowledge with various stakeholders.⁽³⁴⁾ The team included 17 women and men: community members, local authority workers, NGO staff and university graduates. Team members from the various levels of local governance included a woman from the local municipal planning and statistics office, a woman who is a social development officer from the local ward office, and two members from the local sub-ward municipal office: the chairman of the residents, and a woman from the sub-ward committee. There were also three Ardhi University graduates (two men and a woman) who were interns at CCI, and four CCI staff members (two women and two men). The team also included six informal settlement residents, five women and a man, active in the Tanzanian Urban Poor Federation. It should be noted that they have themselves experienced conditions of poverty and unemployment, and are very familiar with the lived experiences of women living in the settlement.

The team began designing the research by discussing and situating the notion of “prosperity” and its translation into Swahili. They decided that it would best be translated as “the good life” – maisha bora. Next, they made collective decisions about sampling and assembling focus groups to discuss situated perceptions of maisha bora in three informal settlements. The allocation of focus groups reflected the research team’s purpose of representing diverse socioeconomic variables, various lived experiences and different household types. It also reflected ethical considerations concerning the expected social dynamics within the focus groups, and the anticipated impact of these on the participants, the researchers and the data collected. In this sense the team members, who later facilitated these discussions, saw themselves as “carers”. To use Tronto’s framework,⁽³⁵⁾ they recognized a need for care (an act of “caring about”), and they debated options for meeting this need.⁽³⁶⁾ This was discussed in relation to the long-term research aim of providing “context-specific policy-relevant knowledge about prosperity”⁽³⁷⁾ through the gathering and analysis of data. But the researchers were also seeking concrete ways to meet the need for care in the short term. The difficulties of considering the impact of short- and long-term actions in tandem led to an ethical debate regarding care giving, which questioned whether the process of focus group member allocation could also be understood as an act of “caring for”.

The use of focus groups in social sciences research, and the ways they may benefit their participants, is widely discussed in the literature. Focus groups are sometimes referred to as “group interviews”,⁽³⁸⁾ which differ from individual interviews because they provide important insights about social interactions.⁽³⁹⁾ As George Kamberelis and Greg Dimitriadis argue, “particular understandings of self are drawn out in focus groups work . . . selves-in-dialogue, social selves, selves-in-community”.⁽⁴⁰⁾ Furthermore, Esther Madriz points out that focus group discussions have a potential for creating a sense of “togetherness”.⁽⁴¹⁾ Yet, the use of focus groups also requires an awareness of its possible negative consequences.⁽⁴²⁾ While minimizing harm is one of the central principles of contemporary institutional research ethics, scholars have stressed the need to move beyond “minimizing harm” to ensure that reciprocal benefits come out of research projects.⁽⁴³⁾

Kamberelis and Dimitriadis theorize focus groups as offering three interrelated key functions – pedagogical, political and empirical:

The pedagogical function foregrounds the dialogic nature of focus group interactions, as well as the possibility for transformative encounters. The political function highlights the sources of collective support that occur around social and political issues. The empirical (or inquiry) function alerts us to the deep epistemological issues and concerns around “the research act,” including the complex negotiations between “self” and “other” in inquiry. ⁽⁴⁴⁾

In the case of this research, the use of focus groups was intended to contribute to all three dimensions. It had a pedagogical function: to build the capacities of the research participants as well as of the team of researchers. It also had political aims: to give a voice to urban informal settlement residents; to strengthen connections and networks among research participants; to build relationships and enable dialogues between urban dwellers and public officials; and to influence urban policies in the long run. To this end, the co-production methodology adopted here was designed strategically to enhance the authority of the findings in the eyes of decision makers. And it also had an empirical aim, in the sense that it planned to involve community members in the collective generation of data through group interactions that would reflect their epistemological perspective.

VI. Debating Focus Group Dynamics

As mentioned above, the first stage of developing a Prosperity Index includes qualitative data collection, aimed at capturing the lived experiences of settlement dwellers, as described in their own terms.⁽⁴⁵⁾ As the research team worked to plan the composition of the focus groups that would be involved in this aspect, they tried to anticipate the dynamics that would take place within them. As also noted earlier, the literature on focus group dynamics provides insights into the complexity of the relations between the self and the social.⁽⁴⁶⁾ A nuanced understanding of group interactions is suggested by Wendy Duggleby, who discerns three levels of data generation: individual, group and group interaction. She notes that interactive patterns within focus groups are in themselves a source of data which is often underused and underreported.⁽⁴⁷⁾ Similarly, Bente Halkier proposes that it is important to integrate various forms of social interaction in analysing focus group data, including both dialogic and non-verbal information, thereby “dissolving the traditional firm distinction between interview-data and observational data”.⁽⁴⁸⁾ Jenny Cameron additionally points out that the interactive aspect of the discussion influences the data, and is therefore inseparable from the social construction of the knowledge produced.⁽⁴⁹⁾ Based on these observations, we consider the process of co-designing the focus groups in Dar es Salaam to be a site of social enactment, in which content and group dynamics were integrated.⁽⁵⁰⁾

Criteria for the formation of focus groups are also debated in the literature. Some scholars suggest that disagreements can be beneficial for the discussion, since, as Barbour points out, “a little bit of argument can go a long way towards teasing out what lies beneath ‘opinions’ and can allow both focus group facilitators and participants to clarify their own and others’ perspectives”.⁽⁵¹⁾ A diversity of opinions is also viewed by Kamberelis and Dimitriadis as highlighting “complexities, nuances, and contradictions with respect to whatever is being studied”.⁽⁵²⁾ Other scholars argue that a common background and similarities between participants can make it easier to share ideas and experiences in a non-judgemental environment.⁽⁵³⁾ Madriz views focus groups as a feminist mode of enquiry, a process of collective testimony:

[Focus groups] are a sensitive tool to recover and use the knowledge acquired from women’s subjective experiences of everyday life. These collective testimonies provide women with the possibility of breaking the wall of silence that has suppressed the expression of their ideas and emotions. ⁽⁵⁴⁾

Similarly, Deborah Thien suggests that focus groups offer possibilities for processes of collective, relational meaning-making as well as a feminist approach to the power relations between the researcher and the focus group participants:

Focus groups have been utilized as a feminist method because they offer a potentially more flexible and less hierarchical arrangement than a one-to-one interview. Feminist arguments in favor of the focus group suggest that the larger number of participants in a focus group allow[s] participants to engage more safely in discussions about matters of concerns [sic] to them.⁽⁵⁵⁾

Further, Kamberelis and Dimitriadis suggest that “[f]ocus groups within feminist traditions have also mitigated the Western tendency to separate thinking and feeling, thus opening up possibilities for reimagining knowledge as distributed, relational, embodied, and sensuous”.⁽⁵⁶⁾ This approach, as well as the positions of Madriz and Thien, relate to groups of women of similar socioeconomic backgrounds. Indeed, Madriz argues that mixing women of different socioeconomic status “would have restricted participation in the discussion because their particular experiences are widely different”.⁽⁵⁷⁾

Yet here we consider the importance of taking an intersectional approach to gender difference. As noted, in the Dar es Salaam research decisions concerning the formation of the focus groups and their composition were made during the preparation workshop. The researchers divided themselves into three sub-groups of four to five participants, each of which discussed the composition of the focus groups. These sub-groups were diverse, and their suggestions reflected different identities and positions of power, which led to some disagreements. These were negotiated first within the sub-groups, and then with the entire team. One of the sub-groups suggested holding separate thematic focus groups of tenants, landlords and entrepreneurs, each composed of both men and women. Another sub-group suggested holding separate focus groups for men and women, with the women’s group including women from various backgrounds: female tenants and landlords, women-headed households, working women and unemployed women. The third sub-group suggested not only holding separate focus groups for women, but also separating employed women (including the self-employed) from those who are unemployed.

These different suggestions for focus group formation and composition were presented to the wider research team, and in this forum it was decided that some participants would benefit from focus group discussions with others who shared some common characteristics. It was therefore agreed to conduct a focus group of young people, a focus group of tenants and a focus group of landlords. In a similar vein, the team also decided to hold a separate women’s focus group. However, a heated debate emerged concerning the question of whether to mix women earning money with unemployed women. All team members viewed unemployment status as undesirable, and as evidence of a lack of agency. Unemployed women were seen as oppressed, either because they were forbidden to work by their husbands, or because, for other reasons, they lacked the required resourcefulness and capabilities. Some of the research team members argued that mixed focus group discussions for women could offer pedagogical tools that would benefit the unemployed women, by providing an opportunity for more than just data collection. They thought that although the focus group questions were not specifically about income generation, mixing the women in discussion might encourage those who were unemployed to try to change their situation. This was understood as an important “side effect” or “impact” of the discussion of maisha bora. Members of the team who saw the focus group interaction as a potentially “transformative encounter”⁽⁵⁸⁾ thought of it as a way of “caring for” the participants, by supporting them in meeting their needs in the short term.

Two of the women researchers, both activists in the Urban Poor Federation, explained that offering challenges to women participants in the context of a group discussion was important, as they would be able to learn from the experience of various other women. One of them told us:

Once you mix them together, it is easier for those who don’t engage in anything to learn from those who engage in something. . . . If one who is maybe doing business, shares her challenges or benefits from the business, this must encourage the one who’s not doing anything. . . . And it’s easier to get different challenges from different people within the same group. ⁽⁵⁹⁾

Asked whether unemployed women would feel uncomfortable in speaking up, she pointed out that it was the responsibility of the focus group convenors to encourage everyone to make their voices heard. Aside from an opportunity for networking and learning, she stressed that the aim of the encounter in the focus group should not be limited to discussions about paid work, but also to empowering the women participants in other ways, such as encouraging them to organize and support each other:

Those who don’t work or engage in anything, because maybe their husbands don’t want them to do anything, they [can] still engage in some activities. Like, the women normally have their [savings] groups which are used to give money to each other. . . . So, if husbands say their women should not engage in business, they are [still] able to engage in those kinds of [activities]. They will just create something, which is not a real business or something, but a group which can make them generate income and do other things. . . . If you put them together, their mind will not be the same. They will see another way of generating income.

Some of the research team members disagreed with this approach, and argued that the experiences of unemployed women were different from those of employed women, and that being part of mixed groups would make them feel inferior. One woman, a resident of the settlement, thought that while challenging unemployed women could be a good idea, at the same time they would not feel confident speaking in a mixed group, so separating them would allow them to speak more freely. Others stressed that separating the women would be more ethical, and also more beneficial, since sharing and analysing their problems together, without the presence of employed women, could lead them to think about ways to address their problems without exposing their vulnerability. They considered this approach to also be more productive for the research, as the discussion and analysis of the structural barriers faced by these women could inform the Prosperity Index. One of them told us:

If you’re having housewives in a separate group, it would be easier for you to get all the detailed information: What are the challenges, … Why are they not [working] like others? Okay, despite the challenges, what do they think? ⁽⁶⁰⁾

This woman also thought that participating in a mixed group would be uncomfortable for the unemployed women, who would feel that they might be open to criticism by more successful women. This difficulty could possibly cause conflicts within the group. Although herself an employed woman, when describing her views on this matter she shifted voice to speak in the first person, as if identifying with the position of the unemployed women:

I can say: “I have nothing, she has something.” How can we have the same language? And the problem is I cannot say my secrets in front of those who are doing good. I’m doing bad. I’d better stay with those who are doing bad, so as I can share what makes me that way. But if I’ll be in a group of those who are doing good, it’s not … it’s really hard!

Although she agreed that the unemployed women could benefit from the advice of employed women, she thought this would best be done indirectly, through the actions of the convenors, who could invite one of the employed women to present her story and in this way offer advice by example to the unemployed women. Here, she spoke as a convenor:

So . . . sometimes later, we can call someone who is from the group that is doing, maybe she or he is an entrepreneur, and maybe that person had the same challenges that these people are facing. Then she can explain and express … Those people can get the experience that, “Oh, I was like you guys, but it was like this …”. So, through the experience and the exposure that she has, those people can learn from her. But they cannot learn when they are in the same group! For them, it will be learning in a very hard way.

In the context of the research in Dar es Salaam, this debate touched upon the unemployed women’s intersectional, marginalized position as both women and non-earning members of the family – those “who don’t engage in anything”.

Following this debate, the research team held a vote that resulted in a decision to hold separate focus groups for unemployed women. This reflected the majority’s understanding that there was a potentially undermining effect in exposing the double challenges faced by the unemployed women. They agreed that potentially these women would feel inferior and refrain from talking, in which case the focus group would fail both to provide a “transformative encounter” for these women in the short term, and to achieve the aim of knowledge co-production to represent a diversity of data – including these women’s lived experiences and views – for the long-term research goals.

The research team agreed that separating the unemployed women would initially allow them to share their experiences among themselves. This would provide data for the research, but it would also support the women in analysing the problems and barriers they faced without the presence of women who are considered more successful. The research team also agreed that, following the separate focus group meeting, they would consider inviting federated women, members of savings groups and businesswomen to meet the unemployed women. Further, the research team intends to hold a follow-up meeting with residents from each of the three communities to present the findings from the focus group discussions. This meeting will provide the team with feedback from the residents, and enable co-learning among members of different groups.

In addition to the focus groups mentioned (mixed-gender groups of young people, tenants and landlords, and separate groups of working women and unemployed women), the research team also decided to conduct a separate focus group of men entrepreneurs, and a mixed-gender focus group of disabled participants.⁽⁶¹⁾

VII. Reflections

The final decisions concerning focus group formation and composition reflect the research team’s collective judgement on the types of groups that would generate outcomes that were, to return to Kamberelis and Dimitriadis, research-oriented, but also pedagogical and political.⁽⁶²⁾

The decision to separate employed and unemployed women points to the team’s conclusion that “caring for” unemployed women should not be done by exposing them to women whose working status could present to them a challenge. Bearing in mind that “caring for” may risk a potentially paternalistic approach,⁽⁶³⁾ it is important to ask whether members of the research team that supported holding a mixed group could have been assuming that they were in a better position to understand the interests of the unemployed women than the women themselves were, or that they were underestimating the agency of these women.

This question should be reviewed in context. As noted, the research team included women who were themselves informal settlement residents. Based on their own lived experiences of extreme poverty and unemployment, they saw the focus groups as an opportunity to support other women. Wilbard Kombe et al. have noted that in Tanzania, about 70 per cent of the urban population live in informal settlements, where poverty is a major chronic problem. They point out that the highly limited opportunities for income generation and employment affect women and youth in particular, thus deepening urban poverty and inequalities.⁽⁶⁴⁾ Over 65 per cent of the urban population rely on informal and often irregular economic activities conducted around their homes, on the streets and in the markets.⁽⁶⁵⁾ Many women living in households with low and irregular income are therefore obliged to find a balance between conservative social norms that require them to perform all household chores, and the necessity to contribute to feeding their families through paid work. In this context, the research team members were aware that vital income-generating activities, or participation in support networks such as savings groups, could significantly contribute to the livelihoods of these women and their families.

Those in favour of mixing the women therefore expected that this experience of social learning would provide the least empowered women with tools and motivation; in other words, this would indicate an act of care that was not simply an act of giving, but that involved an act of receiving. Such caring has to be completed by another,⁽⁶⁶⁾ through their subsequent actions – illustrating, if we return to Tronto’s four categories of care, how “caring for” involves “care receiving” as well as “care giving”. Eva Feder Kittay has also suggested that “[w]ithout the cared-for’s uptake, the caregiver’s actions are not yet care in a fully normative sense”, and therefore that “being cared for, is an active uptaking and not a passive submission”.⁽⁶⁷⁾ In this sense, using the focus groups as a means to encourage women to change their situation is itself an act of care that calls for a response from the women involved, and so respects their agency.

However, this suggestion around how the research activities of focus groups in processes of knowledge co-production might be understood as a relation of “care giving” and “care receiving” also raises questions about the extent to which the focus group experience can be considered to be “challenging” before becoming unacceptable. This question was reflected in the debate on whether “challenging” the unemployed women would benefit them or whether it could cause harm. Regarding the ethical principle of “benefit not harm” and its application in focus group discussions, Julius Sim and Jackie Waterfield point to the difficulty “of distinguishing ‘harm’ from other unpleasant feelings and experiences, and thereby determining whether a moral wrong has been committed”.⁽⁶⁸⁾ In the discussion about the focus group formation, a certain amount of “challenging” was seen as part of an act of “caring for” the unemployed women. But the final decision to hold a separate group for them reflects a preference to remove such a “challenge”. This is because the potential harm was imagined to be worse than merely putting the women in an uncomfortable situation. The concern was that the unemployed women could be silenced or even humiliated, and that this might undermine both the short-term goal of supporting them and building their capacities, and the long-term goal of including their voices in the knowledge co-production process.

We suggest that the solution agreed by the team during the planning phase, following the exchange of many differences of opinion, could offer a way to address both aims, namely to hold separate focus groups for unemployed women first, and then later encourage social learning by inviting working women to share their experiences with the unemployed women and give advice about ways of generating income. This solution was also practical in that it allowed for flexibility, so that the convenors were able to meet the specific participants first, and to then assess the group dynamics and make decisions on subsequent steps. Given that the research was designed to span communities in three informal settlements, such an approach allowed for the knowledge co-production process to be better tailored to different situations on the ground.

VIII. Conclusion

In conclusion, then, it is important to revisit the notion of knowledge co-production as collaboration between partners from different sectors. As we have seen, all the partners in this research agreed on the importance of co-producing a Prosperity Index, a tool for influencing policymaking towards increasing urban equality. The co-production process aimed to allow diverse voices and different forms of knowledge to be heard and recognized, and to open ways for new understandings of the meaning of prosperity in the context of informal settlements in Dar es Salaam.⁽⁶⁹⁾

CCI staff have pointed out that because the words maisha bora were used to describe the focus group discussions, some of the residents who were invited to participate had developed expectations that they might gain some form of material support towards actually achieving “the good life”. As we have seen, although this was not one of the research objectives, the possibility for research participants to gain some practical added value from the focus group discussions was in fact one of the central motivations underlying the debate about the women’s groups. CCI and their local partners had some hopes of ensuring that this project would not only generate academic knowledge on how to address poverty, but would also consider some action plans to initiate small poverty alleviation programmes. These hopes have not been met.

The mismatch between expectations and actualities highlights the different outcomes anticipated by collaborating partners from different sectors. It also raises questions about the separation of research funding from poverty alleviation programmes in donor agencies’ agendas. While this discussion is beyond the scope of this paper, in this international research collaboration it is relevant to the interrelated layers of power asymmetries and inequalities operating at various scales, from global to institutional to local, and specifically how these scales operated between the London-based academic partners, the Dar es Salaam-based NGO, the diverse local research team and the focus group participants. While all stakeholders were interested in using the Prosperity Index to support local communities to influence urban policymaking, the partners in Dar es Salaam expected some immediate actions to take place on the ground as well. They thought that the project’s inability to meet these expectations had left the aspiration of “caring for” incomplete. Indeed this suggests, as we move between Tronto’s categories from “caring about” to the more proactive acts of “caring for”, that the reciprocity of “care giving” and “care receiving” needs careful articulation in relation to expectations.

In this project, the funders and the London-based global partners, who were focused on achieving the long-term project objectives, did not meet the local team’s expectations for immediate support to vulnerable residents through the initiation of poverty alleviation programmes. The local partners considered the global partners’ understanding of “care giving” to be limited by power inequalities that prevented them from seeing the full complexity of the situation. This means that in an international research group it is vital to address the differences of understanding that can emerge concerning what care means.

We start to see how acts of caring operate not only in the four ways outlined by Tronto. In particular, “care giving” is not a fixed category, since it is perceived differently across partners and participants in the processes of knowledge co-production, and in this case, those involved in co-producing the Dar es Salaam Prosperity Index. The multiple interactions between co-producers at various scales involved different ethical issues, from questions concerning the determination of funding priorities to (mis)understandings regarding ideas concerning “giving back”. We argue, then, that despite the value of knowledge co-production for generating transformation in the long run, its contextualization through this specific situation reveals a necessity for simultaneously addressing the differing needs of all participants, not least the urgent needs of intersectionally marginalized participants, in the short term. Indeed, a programme that would include immediate practical measures such as training and seed money could have made it unnecessary to motivate the least empowered women through tips and “success stories” from more dominant working women, which might have the unintended effect of silencing or humiliating them. We conclude by suggesting that ethical awareness of both long- and short-term modes of “caring for” should always be understood in relation to situated acts of care giving and receiving, and in this way could better support “knowledge in action” initiatives for addressing urban inequalities in context.