Methods and research data

establising relationships

In total, I spent a little more than six months in India in three periods from 2004 to 2007. The main location of the field work was Jaipur, the thriving capital of the state of Rajasthan, which has a population of 2,1 million according to the Census of India (2001), or 5,2 mil-lion people according to the official web pages of the Government of Rajasthan (2009). (See Appendix 1 for a map of India). Rajasthan has a population of 56,5 million (2001), and is characterised by high levels of mortality, fertility, morbidity, under-nutrition, illiteracy and social inequality and a slow decline in poverty (Rajagobal 1999, 123), and by a rich history and culture. Although overall poverty rates have declined steadily, Rajasthan has among the lowest human develop-ment indicators in terms of education, health and gender equality in India (UNDP 2002).

My interest in India stems from several visits over the years to In-dia and South Asia in different roles. The reasons for choosing Jaipur as the main site involved several decisions along the way, which were considerably aided by a few individuals and organisations. When I decided to study domestic work in India, I initially thought of focus-

104 See Shah (2000, 112) for an eloquent description of how the domestic workers he interviewed in Nepal asked the breath-taking question: “What will happen to me after you do this writing?”

ing on child workers, so I started the research process by contacting organisations that work on this in India. At the same time, Save the Children Finland contacted me and offered to fund part of my field work costs in return for information sharing, giving some training sessions, and writing a policy report upon the finalisation of my dis-sertation. Given that there were no strings attached to their offer, I accepted it. The contacts provided by Save the Children in Finland and their Indian personnel in India undoubtedly determined my geo-graphical choices, and were very valuable when I began my research.

Save the Children Finland and some of their local partner organisa-tions provided me with some useful initial contacts and helped with practical matters such as finding an apartment. However, I was to-tally independent in terms of research design and process, and the organisations were not aware of my research sites.

During the preparatory field work in 2004 I conducted pilot terviews with workers and employers in Kolkata and Jaipur, and in-terviewed people in several organisations which work with domestic workers in Delhi, Mumbai, Kolkata and Jaipur. These interviews had a double role as they both guided the research design and formed part of the data, although they were not as intensively analysed as the other interviews.

On the basis of the pilot phase I decided to locate my research in Jaipur for several reasons. Initially, I had been thinking of Kolkata and/or Jaipur as research sites since Save the Children UK had pio-neering programmes with child domestic workers in both, and was studying child domestic work in West Bengal. However, I decided to narrow the geographical focus to one city for reasons of quality and efficiency.¹⁰⁵ In addition, others had already studied paid domestic work in Kolkata (Ray 2000b) and a few other cities, but not in Jaipur.

The Institute for Development Studies Jaipur also welcomed me as a

105 See Marcus (1995, 95, 99–100) for a discussion on the challenges of multi-sited ethnography, which has become increasingly popular among anthropologists.

visiting researcher, a mandatory status for obtaining a researcher visa.

Hindi is widely spoken and understood in Jaipur and it was possible to improve my language skills in my home university, whereas Ben-gali is spoken in Kolkata. Finally, Jaipur admittedly seemed a more convenient place for my field work than Kolkata, despite the latter’s hectic beauty, when I had a one-year old child with me.

On my second visit to Jaipur, at the end of 2005, we (I, my then husband, and our daughter) moved into a peaceful middle class neighbourhood where most inhabitants employed domestic workers.

We rented a two-room flat with an attached kitchen from a wealthy business family, who lived downstairs. The family employed two live-in workers whom I saw at work every day, which taught me some-thing of domestic labour practices and relations although I decided to interview neither the proprietors nor the workers for my study.¹⁰⁶ Since it was the task of one of these workers to clean our side of the house as well, unlike Dickey (2000a) or Kumar (2006b), I did not have the opportunity to act out the role of an employer in India. Ku-mar (2006b, 11–12), a native of India residing in the US, felt that it would have been impolite towards Indians not to employ domestic worker(s), which was considered “a necessity just as purified water or air coolers”.¹⁰⁷

Whether having one’s family in the field has a positive or nega-tive impact on the research depends on many factors, particularly the topic of the research (see Burns McGrath 1998, 64–65; Young Leslie 1998, Linneken 1998; Ronkainen 1989).¹⁰⁸ My topic was so

106 Eskola & Suoranta (2000, 99) argue that there is sometimes a thin line be-tween spying and observation. On occasions, I also got to see intimate aspects of relations between the house owners and their workers, and could not help feeling that this was unintentional spying.

107 See also Kidder (2000) for her experiences as a foreign employer of domes-tic workers in India.

108 For Burns McGrath (1998, 64–65), studying the decision-making process regarding illness and dying, being a mother and having her three children and husband in Tonga with her, significantly eased entering into informal conversa-tion about the topic of study.

integrally linked to mundane, domestic daily routines that visiting the local homes as a married woman and a mother probably made it easier to approach the subject. My experiences during my research were similar to those of others who felt that being married or being a mother legitimated their work and put them in a superior posi-tion to an unmarried or childless woman, a situaposi-tion I myself had experienced on previous visits to South Asia. (see Sinclair 1998, 127;

Linneken 1998, 79; Latvala 2006). ¹⁰⁹

Having my (exotic) daughter with me in India helped to pave the way when getting to know people and finding participants for the study in a relatively short time (see Latvala 2006, 44; Kontinen 2006, 236). Walking around the quiet streets of the middle class neigh-bourhood and watching dogs, cows, birds, flowers and people was the favourite pass-time of my daughter who had just learnt to walk.

The only others to walk the streets were street vendors and domestic workers. The residents didn’t seem to mind this departure from Indi-an middle class conformity: on the contrary, they were very tolerIndi-ant of such working class habits, given that I was an outsider and clearly an educated woman who could afford to rent a relatively expensive flat.

I (“the mother of Molla”) and my daughter were constantly invited to visit the homes of both familiar and unfamiliar neighbours, for the ubiquitous tea and biscuits.¹¹⁰ As a result I could start having informal discussions about the topic right after we settled in, and the interviews in this neighbourhood paved the way for the employer interviews in other locations, although we lived in this location for only the three winter months of 2005–06 (I stayed elsewhere in 2004 and 2007).

109 Flinn (1998, 109) writes about how the presence of her family in the field made her visibly a more complex person with a variety of roles, which were also under close observation.

110 Flinn (1998, 104) found that for the Pollapese people in Samoa she was often remembered of or referred to as being “the mother of Colin”, “the wife of the American teacher, Jim” and so on. First, she had become somewhat disconcerted at being so invisible or forgettable but with time came to interpret it as a measure of a certain type of success, having possibly been appropriately unobtrusive.

observation and interviews

In order to understand domestic labour relations better, I used sev-eral “tracking” strategies (Marcus 1995, 95). I read extensively aca-demic literature, policy reports, and media articles on domestic work in India and I talked to people in different positions, experts and non-expert Indians. However, the main methods of my work were the qualitative interview and observation.

My study is ethnographically oriented in that I aim to get a holis-tic picture of the phenomenon that I study (Eskola & Suoranta 2000, 105). There were elements of participatory observation such as daily socialising with some of the employers and attending social events in one of the neighbourhoods even if my observation was not par-ticipatory in a strict sense (Grönfors 1982, 87-88).¹¹¹ The decision to spend three shorter periods in India instead of one longer period was partly dictated by pragmatic questions. The relatively short dura-tion of each visit may have limited my chances for understanding the micro-politics of the research situation (Reger 2001). However, the repeated visits and interviews with the same persons in 2005-06 and 2007 proved very helpful. Having been able to read the data carefully in-between the field work periods and having been able to read it

‘against’ other studies on domestic work, I was able to see things in a new perspective and notice issues that I had not during the previous visit (see Eskola & Suoranta 2000, 102). Visits in consecutive years also made it possible to observe time-related changes in domestic la-bour relations, and to see how life course and work intertwine.

Since I was interested in how employers and workers themselves perceive domestic labour relations, and in why they act the way they do, the qualitative interview was a suitable method for this study (Tuomi & Sarajärvi 2002, 74). This method also suited my aim to provide a detailed, contextualised and situated description of labour

111 I have been able to passively observe domestic work practices during previ-ous visits to India, Bangladesh and Nepal from the early 1990s onwards.

relations (Weiss 1994, 9). Since my main interest was not on quanti-tative information but on human relations – more specifically labour relations – a survey method would not have been suitable. In addi-tion, half of the interviewees were illiterate, which would have com-plicated the process and, possibly, led to misunderstandings (ibid, 75). Besides, the Institute of Development Studies in Jaipur was simultaneously conducting a survey on domestic workers in Jaipur commissioned by the ILO, so there was no point in repeating this.

Although the scope and methodology of that study is different, it served as a useful reflection point and helped to validate some of my findings.

I carried out semi-structured, thematic inteviews with twenty-one domestic workers and seventeen employers in Jaipur, and talked less formally with several others.¹¹² All names of the workers and the employers have been changed to protect their privacy. I interviewed most of them twice in consecutive years, and some of the workers three times. The first interview with each participant was relatively structured and followed more or less the same course. Subsequent interviews had certain common themes but a more open structure, which allowed for more dialogue and variation in the order of themes.

(see Tuomi & Sarajärvi 2002, 77).

Interviews with all workers and with nine employers were made in Hindi, and with eight employers in English. All interviews were recorded and transcribed, the English interviews mainly by myself, and the Hindi interviews both interpreted and transcribed by Gargi Gopesh, a native speaker of Hindi. She also understood Bengali, the mother tongue of some workers, even though they conversed in Hindi. While the language did not become a major barrier, the scope for misinterpretation posed by the use of interpreter can never be fully mitigated.¹¹³ Although my Hindi skills were limited, it was

112 For a list of persons interviewed see Appendix 2.

113 See Lindberg (2001) for her experiences in using an interpreter in Kerala.

an advantage that I knew enough to be able to follow the conversa-tion and comment on possible misunderstandings. Interviews with those employers who insisted on speaking English in spite of their limited skills were another challenge. The pilot interviews carried out in Kolkata in 2004 with six workers and two employers were made in English and in Bengali, interpreted and transcribed by Dr. Kakali Das, who also acted as my research assistant.

Gargi, my invaluable assistant and interpreter was a young mar-ried Brahmin Hindu woman, a mother of a small child. She had a MA degree from an Indian university, and was pursuing studies for a doctorate degree. However, as she had a three-year old daughter whom she took care of at home, she was not studying or working full time during my field work periods. One should note here that there are very limited chances in India to get a position as a funded post-graduate student. Therefore, such students may be willing take jobs clearly below their merits, including as research assistants for foreign researchers who are able to pay for them. Gargi also welcomed the op-portunity to be involved in occasional paid work through my research while her mother-in-law and a live-in worker took care of her daugh-ter during our working sessions. For the respondents, Gargi clearly belonged to the educated middle class. The employers, in particular, usually asked her a few questions which would reveal her caste sta-tus.¹¹⁴ I aimed at mitigating the potential bias of her caste and class status by following a similar structure in all the initial interviews, but it was virtually impossible to mitigate the consequences of her status entirely.¹¹⁵ We discussed the ethical choices and issues several times, but we still sometimes differed in our approaches, as any two people usually do.

114 Srinivasan (1997, 91) found that in her field research in rural India one of the very first questions the interviewees asked her was her caste affiliation.

115 Lindberg (2001, 70–71) felt that having a Christian interpreter in Kerala mitigated the biases which might occur when Hindu women of low caste meet with high caste women.

For ethical reasons, the employers and workers I interviewed were not only from different households but also from different geograph-ical locations. Froystad (2003, 90-91) interviewed workers and em-ployers from within the same houses and felt this led to an overempha-sis of the employers’ perceptions in her otherwise rich ethnographic analysis. The workers she approached were unwilling to talk to her, even when she approached them in privacy. During the pilot phase, I also conducted one interview with three part-time workers in the presence of a woman for whom one of them worked. Even though it was reasonable to expect the employing woman, a long-time human rights activist, to be a caring employer, I think both I and the work-ers felt uncomfortable in her presence. Froystad’s (2003) experiences and the pilot interview supported my decision to rule out interviews of workers in the presence of employers, as well as the interviews with current live-in workers. Interviews with live-in workers, even if done in seeming privacy, could have influenced the content, put both sides in an uncomfortable position, and potentially risked the work-ers’ safety, given the verbal, mental and physical abuse sometimes in-volved in domestic work. Moreover, even though some of the live-in workers visit their own home every second week, I felt it would have been intrusive to ask them to be interviewed during the few hours of free time with their families. In sum, while I acknowledge that close observation of workers and employers within the same house could have revealed relevant issues about the dynamics of their relation-ship¹¹⁶, I feel the decision was the right one.

Since I decided not to interview current live-in workers, excepting only the pilot interviews in Kolkata¹¹⁷, I had to find other ways of acquiring information on live-in relations from the worker perspec-

116 For example, by examining closely workers and employers within the same houses in Indonesia, Kathleen M. Adams (2000, 158) was able to understand the role of humour in negotiations and bargaining in labour relations.

117 Interviewing them was more feasible since they (with permission from their employers) participated in an NGO programme, and the interviews were carried out in the centre, outside the employers’ premises.

tive. The question was solved, partially at least, since seven workers in my study had previously worked as live-in workers, and they told me about their experiences. I also talked with other live-in workers informally.

I found the respondents to my study through a snow-ball tech-nique. In the case of workers, I was considerably aided by an activist woman whom I met through researchers at the University of Rajas-than. She initially put me in touch with two different communities of domestic workers situated near her home. Unlike most educated people in Jaipur, she knew many domestic workers personally but did not employ any. With her continuous support, it was relatively easy to “find” workers, and even to trace some who moved to a new place on my last stay in Jaipur.

Based on prior information about exploitation of workers, I had initially been sceptical of whether I would find employers willing to discuss their relations with, and practices towards, the workers. My concern was unnecessary: all middle class women and men whom I asked to be interviewed agreed, and as mentioned in Chapter 1, talked quite openly about their practices.

In my experience it is not only the researcher who aims to make the interview situation as comfortable as possible, but also the re-spondents. The girls whom I interviewed were particularly caring to-wards me from the beginning. One way to do this was small -talk, for example, about what I and the interpreter were wearing. Although their comments on our suits and on whether the colour suited me or not¹¹⁸ also reflected the significance of what one wears in India (Tiengtrakul 2006, 31),¹¹⁹ I perceived these and other questions as the girls’ efforts to make the situation comfortable for me. In the case

118 See Mookherjee (2001, 4–5) for similar field study experiences in India.

119 Linneken (1998) talks about impression management in field work situa-tions, part of which is the question of what one wears. Since my dress was always more or less the same; a salwar kameez (a long tunic shirt and trousers) with a dupatta scarf, there was no difference in the way I dressed “for” the workers and employers.

of the workers, my initial feeling that I was an intruder who was steal-ing their precious time between two work shifts was also eased by the relaxed attitude of the girls and women themselves to the interviews.

The same was true for some of the employers, who seemed to be making an effort to make me feel welcome and comfortable.

Employers

The ages of the fifteen female and two male employers in my data ranged from around twenty-five to around sixty-five years. As one criterion, I included employers of both live-in and part-time workers in the study. The employers lived mainly in three different Hindu-dominated middle class neighbourhoods in Jaipur. In the state of

The ages of the fifteen female and two male employers in my data ranged from around twenty-five to around sixty-five years. As one criterion, I included employers of both live-in and part-time workers in the study. The employers lived mainly in three different Hindu-dominated middle class neighbourhoods in Jaipur. In the state of