Ethnographic challenges to studying the poor in and from the global South

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ Ethnographic challenges to studying the poor in and from the global South © Authors, 2020 Published version Stark, Laura Stark, L. (2020). Ethnographic challenges to studying the poor in and from the global South. In T. Lähdesmäki, E. Koskinen-Koivisto, V. L. Čeginskas, & A.-K. Koistinen (Eds.), Challenges and Solutions in Ethnographic Research : Ethnography with a Twist (pp. 131-145). Routledge. https://doi.org/10.4324/9780429355608-9 2020 9 ETHNOGRAPHIC CHALLENGES TO STUDYING THE POOR IN AND FROM THE GLOBAL SOUTH Laura Stark Introduction Contrary to what the UN and World Bank have proclaimed in recent years, global poverty has increased dramatically (Hickel 2016, 7). The United Nations and World Bank measure only how many persons are living under an arbitrarily chosen boundary of 1.9 USD per day, but if the poverty line were raised to the minimum for a normal life expectancy (approx. 4 USD in most countries) then 3.5 billion persons –half the world’s population –would be classified as poor (Edward 2006; Hickel 2016, 7). Anthropologists have long studied people in low-income countries typically defined by poverty, but have rarely brought their ethnographic methods to bear on the causes and consequences of poverty per se (Ferguson 1997; Booth et al. 1999; Green 2006). The focus of anthropology in the Anglophone literature, for instance, has tended to be rural areas and the study of one ethnic group rather than ethnically diverse urban areas. Topics tend to derive from anthropology’s“primitivist reflex”(Kalb 2015, 52) that is, the search for cultural and social aspects with minimal influence from colonialism, market capitalism, globalism, or mass consumption. There has been much less interest in using ethnography to study how poverty affects the people studied through ethnography, including asylum seekers and migrants coming from low-income or materially inadequate conditions. Yet ethnographic research on poverty is sorely needed. The micro-economic research dominating studies on poverty lacks the necessary tools to study socio-cultural power dynamics, and often misrecognizes “on the ground”complexities in the lives of the poorest (Booth et al. 1999; Bevan 2004, 29; Ferguson 2015). My aim in this chapter is to encourage ethnographers to carve out a space for themselves in research on poverty in the global South by recognizing their strengths in this area, namely sensitive methods in local contexts to answer questions of why something happens rather than just how much can be measured. Yet qualitative researchers confront very different socioeconomic dynamics in the global South than they do in the global North. They therefore need to be aware of ethical and methodological challenges in studying the poor in the global South. In this chapter I discuss the following challenges: 1) understanding the expectations of persons coming from societies where dependence networks structure socio-economic relations; 2) informed consent among the poor in the global South; 3) personal data among those with meagre identity documentation; and 4) the effects of poverty and stress on participants’memories. These issues call for not merely new methods, but a new overall approach to ethnography that understands how poverty affects the information given by those studied. More broadly, I propose that Northern ethnographers need to rethink familiar ethical and methodological approaches in a “twist”towards a “Southern”approach that encompasses both reflexive recognition of socio-economic differences and the alternative methods of perceived causation and third-person elicitation. These methods allow the qualitative researcher to avoid asking for personal data and to circumvent the non-verifiability of individual life facts. The term ethnography in this chapter refers to the presence of the researcher in the everyday spaces in which participants live. It also indicates an “open notebook” approach in which interview questions are not predetermined by a research design but instead are adapted to the new information received from participants as the study proceeds. This approach ensures flexibility in asking new questions and the discovery of things not imagined to exist when the research began. When used among the urban poor, the open notebook approach allows researchers to see everyday life and social dynamics in ways they could not have envisioned based on their own non-poverty backgrounds. These moments of discovery serve as clues leading to new and interesting paths of inquiry. The setting of my research As part of an ongoing study of gender and urban poverty carried out between 2010 and 2018, I interviewed a total of 292 persons in two low-income, predominantly Muslim neighbourhoods in Dar es Salaam, a city of 4.36 million inhabitants (National Bureau of Statistics Tanzania 2013). Here, an estimated 50% of the population in informal settlements live on an average income of roughly 1 USD per day, well below the international poverty line of 1.9 USD (Ndezi, 2009). The majority of those I interviewed (roughly 70%) had only a primary education or less. The neighbourhoods I studied fulfilled UN-Habitat’s (2010, 14–15) criteria for “slums”as they lacked secure tenure, sufficient living space, sanitation infrastructure, easy access to safe drinking water, and durability of dwelling structures. As a port city that draws migrants from throughout Tanzania, Dar es Salaam is an ethnically heterogeneous but unilingual Kiswahili environment. Although no official census data exists on religious affiliation in Tanzania, Sunni Muslims comprise roughly 75–90% of the residents in the neighbourhoods I studied. Interviews were 132 L. Stark conducted in Kiswahili with the assistance of female interpreters and female key informants who had first-hand familiarity with the socio-economic circumstances of those interviewed. Key informants and interpreters were found through local non-governmental organizations. All names of interview participants in this chapter have been changed to protect personal anonymity. I discovered soon after I had begun interviews that I was extremely fortunate in the random choice of my first field site: I had unintentionally begun interviews in a locality that was sympathetic to outsiders for reasons linked to local and national politics. The local authorities there welcomed me immediately, which set the tone for other persons in the neighbourhood. In the neighbourhood just across the road, however, I was not so lucky: a different political situation there meant that the local authorities viewed me with suspicion and demanded a different bureaucratic process in order to grant permission for interviews. Although I conducted some interviews there, I eventually focused fully on the first neighbourhood, which was home to roughly 2,500 persons. Skype as an alternative to face-to-face interviews In 2018, health issues prevented me from traveling to Tanzania, so I decided to conduct my final round of interviews using Skype. Over the last decade, there has been growing attention to alternatives to traditional face-to-face interviews in qualitative research such as Voice over Internet Protocol (VoIP) technologies (Sullivan 2012; Lo Iacono et al. 2016; AlKhateeb 2018). Most existing literature on the use of Skype describes interviews conducted among middleand higherincome participants. Thanks to the dramatic uptake in usage of mobile phones among low-income persons in the global South, Skype interviews are also possible across vast distances and with chronically poor interview participants. Through emails with Zakia, a female interpreter with whom I had worked twice before, I explored this possibility. Zakia downloaded the Skype program onto a basic smartphone that I sent her money to buy. She then contacted Neema, a woman in her late 40s who was our key informant in the neighbourhood. Neema agreed to ask residents to participate in interviews as she had done in previous years, and Zakia travelled to the neighbourhood at the agreed time. The strength of our Skype connection was not sufficient for continuous video, but we were able to conduct the interviews using audio. Whereas in previous field visits I had always begun with a visit to the local government office to obtain their permission, now Zakia went on my behalf to ask the government secretary permission to conduct the interviews via Skype. This permission was granted, and I conducted 20 Skype interviews. Each interview lasted longer than it would have normally taken face-to-face due to the poor quality of the audio and the need to repeat questions and answers. Although challenging (often I had to nearly shout my questions into my laptop to be heard by my interpreter), I obtained important new data for my research. Interviewing without video was surprisingly easy, given that I was already familiar with the flow Ethnographic challenges in the South 133 and rhythm of the interview situation in Swahili. Both Zakia and interview participants were highly accustomed to the use of mobile phones, so this helped to make interviews through mobile-enabled Skype flow smoothly. Zakia also worked hard to ensure comprehension by enunciating clearly, turning to speak directly into her phone when interpreting for me, waiting for me to ask questions, and patiently repeating things. However, when compared to face-to-face interviews, I noted significant differences. It was frustrating to not observe what was happening in the real-life conversation area, for example, I could not see gestures or facial expressions, and I could not immediately see who had newly arrived to join in the conversation. Skype interviews also exacerbated the ethical dilemma of informed consent. Zakia always told interviewees the purpose of the research and asked for their consent before the Skype call began, but I could not know how the participant had reacted. One benefit of using Skype for interviewing was that having only two Skype interviews per week rather than face-to-face interviews every day as I had done in the field provided valuable time needed for reflection before the next interview, time that I had rarely enjoyed when in the field. Skype interviews also created a psychological distance between the participant and myself. Many of the stories told by interviewees were distressing, for instance, when HIV sufferers and their relatives told of their suffering, or when sex workers told of the violence and abuse they had endured. Face to face, I had experienced such narratives as important but emotionally exhausting. It was considerably less stressful to listen to them at a distance without a visual connection. What I experienced as sensitive or disturbing topics for discussion, however, were not necessarily what interviewees seemed to experience as oppressive. In particular, experiences of not having relatives to help or support financially seemed emotionally difficult for some persons to talk about. Skype interviews would not have been useful as a means of gathering initial data in another country if I had not spent considerable time there already. A relationship of trust with key persons in the neighbourhood as well as with an interpreter were necessary for the success of these interviews. Once relationships of trust with local persons are already established, however, long-distance Skype can create an opportunity to collect additional data if travel is not feasible. Dependence networks I visited fieldsites only during the day and was accompanied by at least one other person as I walked through the neighbourhood: usually my female interpreter or female key informant. I never felt that I was personally in danger, but the fact that I was so easily identifiable as a foreigner by my skin colour meant that I could not live or sleep in the neighbourhoods I studied: I would have been an instant target of theft, since many residents assumed that any white person or mzungu was extremely wealthy. Indeed, just by arriving in Tanzania I had shown that I had enough money to fly from another continent, a feat that most residents assumed would always be beyond their means. Everyone assumed that I possessed more 134 L. Stark money than I would ever need, therefore nearly every meeting with a Tanzanian local became a negotiation in which I was asked for financial assistance: to buy a house for a taxi driver, to help pay for children’s school fees, or merely to “help” in the form of giving cash. This quickly became exhausting, for although I understood people’s motives for asking, it was difficult to create friendships when people (rightly) perceived that my own economic situation was so different from their own. Later in my fieldwork, my desire to give small amounts of cash to persons in extreme need was frustrated by the fact that I was continually short of cash because the local ATM machines were either empty or difficult to reach from my hotel on the city outskirts. In cities where all transactions are over-monetarized cash is vital for survival but difficult for the poorest to obtain (de Sardan 1999). People were unable to grow enough food for themselves in an urban neighbourhood, and needed money to pay transportation, rent, medicines, medical bills, and school fees for their children. The end of interviews were often moments when some informants asked for monetary assistance. Some persons told me in repeat interviews that I had promised last time to bring them a gift, although I had no recollection of doing so. At the time, I experienced these situations as uncomfortable, because I did not want to “bribe”my informants with money or gifts, and I did not enjoy being pressured. It was clear to me that everyone needed help desperately, and impression was often overwhelming. At first, drawing a firm line against giving money was a way to mentally block out the distress and suffering all around me during interview visits. I was also worried about the inflationary effect of monetary help on interview participants. If one resident heard that someone else had received help, they themselves would expect help, and possibly a larger sum of money. For this reason, I gave each participant a tin of powdered chocolate drink (that would have been expensive for them in the shops). A PhD student who conducted interviews separately in the same neighbourhood later told me that participants had complained to her that this chocolate drink was useless for their every needs. One interpreter suggested that sugar (for tea-drinking British style) might make a more useful gift, and so I began to give each participant a two-kilogramme bag of sugar. Yet this did not help those persons who had nothing to eat, no food to give their children, or no way to pay their rent. Upon reflection, knowing that just a few euros would make a difference in many person’s lives, I began making exceptions to my no-money rule for single mothers with no income, persons infected with HIV, and recent migrants to the city sleeping in a relative’s house who had no income. I had not yet realized, however, that when interview participants asked for “help”they were actually striving to create a relationship of dependence upon me. What they wanted from me was not simply money, but a relationship that would confer upon them a sense of security as well as the social identity of being the friend of an mzungu, a white person imagined to be wealthy. Once I became more intertwined in a several participants’networks of assistance, I began to reflect further on how these networks were, in fact, structuring nearly all Ethnographic challenges in the South 135 of the socio-economic relations within the neighbourhoods studied. Anthropologist James Ferguson (2013) points out that in much of sub-Saharan Africa relations of dependence have been preferred or prioritized over independence. In Africa, people rather than land have historically been the source of wealth, so the goal has often been to build hierarchical relations in which persons compete for followers. Both patrons and dependents have therefore benefited from “wealth in people”(Ferguson 2013; see also Bledsoe 1980; Iliffe 1987; de Sardan 1999), a term still used by Tanzanians today. Regardless of where one fit in the social hierarchy, dependence conferred social identity: “without networks of dependents you were nobody […] with them you were a person of consequence”(Ferguson 2013, 226). As the African continent is marginalized within the global network of trade and many unemployed are now locked out of the labour market, what used to be ‘kin networks’in the anthropological literature are no longer necessarily robust or numerous social linkages, but often rely on a few persons, whose support has become meagre, irregular, and unreliable (Cleaver 2005). Social networks have shed dependents, and unclaimed persons scramble to find new persons to be dependent upon (Cleaver 2005; Ferguson 2013; 2015). Advance awareness of participants’socio-economic expectations can ease an ethnographer’s entrance into the field and make it easier to draw clear personal boundaries. An awareness that power, income and social relations in the global South are frequently organized through dependence networks also has implications for ethnographers’work with migrants and refugees. As has been widely documented in the migration literature (most notably in work on remittances), dependence networks may motivate or obligate migration, and often continue after an individual or family has migrated. This means that migrants can depend on already existing ethnic, religious and/or kinship networks in the country to which they migrate. They may also be under heavy pressure to send money to kin or friends in the departure country or in refugee camps, which limits the amount they can invest in their own socio-economic integration in the new country (Lindley 2010; Hammond 2011). They may, for instance, find themselves forced to engage in lowpaying and exploitative work rather than continuing their education because the risk of losing income is too great when they need to send money regularly (Humphries et al. 2009; Eversole and Johnson 2014). Asylum seekers may also expect to establish long-term relations of dependence with the volunteers and reception centre workers whom they see regularly (World Relief 2017; EURITA 2019). Informed consent Relationships of dependence in the neighbourhood were linked to another methodological and ethical challenge I faced while interviewing: informed consent. To obtain valid free, rational and informed consent from research participants, participants must be accurately informed of the purpose, methods, risks, and benefits of the research, and must understand this information (Campbell 2009). This notion of informed consent, however, rests on Western understandings of the person as an individual. Although I received verbal consent from all participants, the persons whom I interviewed had no way of making “independent”or “free”choices for several 136 L. Stark reasons. They did not have the educational background taken for granted in Europe or North America that would have allowed them to be able to evaluate their best interests with respect to the interviews I was conducting. Some of my interview participants did not know what a university was, much less what basic research was. Despite my best efforts to explain myself and the purpose of my research, many participants remained convinced that I worked for an NGO, which was for them the most familiar role for a white woman in Dar es Salaam. Many of them were dependent upon social relationships of neighbourliness that revolved around my key informant Neema, who was a respected and trusted middle-aged female member of the community. It is possible that persons she invited to participate in interviews found it difficult to refuse her. Finally, many participants were so desperate financially that they might have been reluctant to refuse the bag of sugar that I gave to everyone who was interviewed. Overall, there exist very few options for ethnographers to ensure that the people studied participate voluntarily, with no strings attached. Ethnographers can only strive to recognize how Northern assumptions of “free”participation are unsettled by local circumstances, and describe this unsettling for reading and listening audiences. The alternative would be to not study the Southern poor at all. This would mean that those whose voices are already stifled by lack of social power and influence would be further ignored and marginalized by social science. Personal data and sparse identity documentation In Europe, the issue of personal data has come to the fore with the European Union’s General Data Protection Regulation 2016/679 (GDPR), which came into effect in 2018. The GDPR strictly regulates data protection and privacy for all individual citizens of the European Union (EU) and the European Economic Area (EEA), as well as the transfer of personal data outside these areas. The term “personal data”within ethnography refers to all potentially identifying markers: personal names, places, exact ages, exact dates, titles or unusual occupations. Ensuring adherence to this regulation when processing and storing any personal data may demand considerable efforts from ethnographers. When studying life histories or personal case studies, it is difficult to avoid collecting such information. However, answering why certain processes or events occur in groups or communities does not necessarily require personal data to be collected, as I discuss further below. Ethnographers taking this approach may choose to collect no personal data at all, and instead record information in the form of fully anonymized conversations and observations (field notes). This means no audio, video or photographic recording, as it is nearly impossible to remove identifying information from these data storage media. Non-identifying information can include gender, approximate age (specifying only 20s, 30s, or 40s, for instance), general education level, and marital status. According to GDPR rules, it is permitted for these fully anonymized fieldnotes to be imported from non-EU (third) countries into the EU without further protective measures. Fully anonymized data has one advantage when studying the chronically poor who may not be literate. Since the principles of data protection do not apply to Ethnographic challenges in the South 137 anonymous information, signed consent forms are not required if no personal data is collected. In my own research I found that in spite of mandatory primary schooling in Tanzania, a surprising number of interview participants reported not having attended school for various reasons and they could not read, much less write. They usually wished to hide this fact from their neighbours, however, and it is possible that many more persons never revealed their lack of literacy. I obtained verbal rather than written consent in order to avoid humiliating these participants, especially since they often came to be interviewed accompanied by neighbours. Researchers conducting longitudinal studies, however, will need identifying data to keep track of the same participants over time. In this section, I focus on the particular challenges surrounding personal data collection in interviews with the poor in the global South. For instance, most researchers conducting fieldwork assume that they can ask an informant’s name, age or other relevant personal data and will receive answers that can be verified –at least in theory –through fact-checking. But this assumption is rooted in the experience of living in the global North, where individuals’lives tend to be highly documented. Northern citizens are constantly reminded by governmental institutions (schools, tax offices, registrars), financial institutions (banks, credit cards) and places of employment that any identity details they tell about themselves should conform to this documented information. When I began fieldwork, I was unprepared for a system of personal data referencing that differed from my own experience. I began each interview by asking the participant’s name, age, educational status, how many children the participant had, and ethnic group affiliation. I assumed that there was only one right answer to each of these questions. I asked about ethnic self-identification because I wanted to demonstrate the ethnically mixed nature of the neighbourhood I studied. Eventually ethnicity and number of children turned out to be the only reliable personal identifiers when I later tried to find someone I had interviewed previously, because names, ages, and educational status could change from interview to interview. I had failed in the beginning to realize that the names told to me by participants could be of three different types: personal name plus family name, as in Europe; or personal name plus father’s personal name. Alternatively, names could consist of a personal name, the father’s name, and the grandfather’s name. But if the participant was a married woman with children, usually none of these names were used or even known by their neighbours. Instead, neighbours knew adult women by the name of their first child, for example, “Mama Hassan”or “Mama Asha”. Whenever I asked my key female informant whether a previously interviewed person was still living in the neighbourhood, she often had no idea who I was talking about, since none of the married female residents knew each other by the combination of names they had told me during the interview. A particularly perplexing moment came whenIbegantorealizethatmanyinterview participants, even young ones, did not know exactly how old they were. Sometimes older informants (over 40) did not know their age or said they were born in a year that did not match the age they had previously told. Sometimes their only way to reckon their year of birth was an important national event that a parent had told them had happened near their year of birth. Additionally, even younger persons occasionally gave different 138 L. Stark Stark, L. 2019. “Motives and agency in forced marriage among the urban poor in Tanzania”. Ethnos: Journal of Anthropology. https://doi.org/10.1080/00141844.2018.1559214 Sullivan, J. 2012. “Skype: An appropriate method of data collection for qualitative interviews?”The Hilltop Review 6 (1): 54–60. http://scholarworks.wmich.edu/hilltopreview/ vol6/iss1/10 (accessed 28 January 2020). UN-Habitat. 2010 (Revised and updated version of 2003 report). The Challenge of Slums. London and Sterling, VA: Earthscan Publications Ltd. World Relief. 2017. “Boundaries prevent burnout”. https://worldreliefdurham.org/esta blishing-boundaries-and-avoiding-burnout (accessed 28 January 2020). Yle Kotimaa. 2015. “Keksityt kertomukset ovat erikoisalaamme: Turvapaikanhakijoille kaupataan mukaan nyyhkytarinoita”. Last updated 5 December 2015. https://yle.fi/uutiset/keksi tyt_kertomukset_ovat_erikoisalaamme__turvapaikanhakijoille_kaupataan_mukaan_nyyhkyt arinoita/8501625 (accessed 28 January 2020). Ethnographic challenges in the South 145