Covid-19 | How moving (academic) conferences online could help address social injustices

Covid-19 | How moving (academic) conferences online could help address social injustices

Curtailing the movement of people around the world in a bid to control and eventually stop the spread of Covid-19 has forced many, including academics, to gather online. A recent ...

Power to the People? The Right to Information Law in Morocco

Power to the People? The Right to Information Law in Morocco

Morocco’s recently enacted Right to Information Law is a potentially powerful tool in the hands of its citizens, but their ability to use it is still largely dependent on the ...

Disasters, Dilemmas and Decisions: Notes from a monsoon fieldwork in Assam, India

Taking an ethnographic route to study disaster-affected communities makes us grow deeply aware of seething worldly inequalities that disasters bring forth. At the same time, it makes us compassionate towards the world outside. It is imperative we reserve a piece of that compassion for our own selves, too, writes Mausumi Chetia.

On a summer night eleven monsoons ago, sleep evaded me. Outside, the winds were growing stronger by the hour and the rain refused to stop pouring. My sleepless thoughts held the image of a family and an incessantly shaking, Kare Okum (chang-ghar in Assamese) or stilt houses built over a flowing channel of water. I was in Majuli, a densely populated island in the Brahmaputra river of Assam, my home state in northeast India. I was collecting data for my Master’s dissertation around the time the monsoons began, bringing the annual floods to our state.

A few days later, my then research supervisor pulled me out of the field site. With a calm, but commanding voice, she asked me to return to ‘safety’ at the earliest. With partially collected data, mixed emotions and a river that was continuously expanding (due to massive erosion of its banks), I left Majuli for my home’s ‘safety’.

Fast forward to fieldwork for my current research in Assam in 2019. I was faced once again with the ethical dilemma of applying a methodological approach focusing on people first and foremost vs prioritising my own safety as a researcher. They were the very puzzles I had left behind in Majuli when I left a decade ago. Research is a deeply political process, but it is a humanising journey, too, as I would come to see (and as Kikon 2019 explains).

The half-written story of a monsoon auto ride: taking a submerged path

My fieldwork in June 2019 took me to one of Assam’s severely flood-prone districts. The geographical location was chosen based on longstanding professional relationships and empirical familiarity with that region.

Map of Assam
Figure 1: Map of Assam, India

A local humanitarian organisation aided me in accessing the research site. I found an accommodation in the district town, about 45 kilometres away from the actual site of study. My research populations lived about 15 to 25 kilometres away from a national highway, on a rehabilitated government-owned piece of land situated along the banks of a river.

The initial fieldwork days were about establishing access and meeting key contacts. The weather was mostly cloudy with occasional showers. One day looked particularly promising, with the sun high up in a clear, blue morning sky. Predictably, the sky got dark in no time. Despite warnings of rising water levels, our auto-rickshaw driver decided to risk continuing the trip. Then the drizzling started.

I could see a blurring silver line across the trees. It was the water. My heart skipped a beat as I remembered the last time I was in a heavily flooded river in a steamer boat, almost two monsoons past. Some thatched houses appeared inundated already. Buffaloes and cows were clutching at tiny islands that had formed in the paddy fields, which in turn resembled huge lakes on either side of the road.

Stranded animals Assam
Figure 2: Stranded animals on islands of submerged paddy fields (photograph taken by the author, July 10th, 2019).

Gradually, it felt like we were surrounded by a sea of water. There came a point where the road ahead looked completely submerged. Our auto-rickshaw came to a halt. A few vehicles had stopped ahead of us. The rain continued pouring. There was chaos and confusion. People needed to cross the submerged part of the road to reach their homes. But the prospect of crossing the waist-deep water with their luggage, infants and children delayed their decision. Three of my fellow passengers told me that they were indeed afraid of the water. Nonetheless, they had to cross it on foot. Devoid of alternatives, two women and a teenaged boy started marching ahead. Many, like them, were from villages as far ahead as 15 kilometres.

An ethical-methodological dilemma

I continued standing there next to our auto-rickshaw, almost in a stupor. A billion thoughts crossed my mind. And here was my dilemma. By design, understanding the ‘everydayness’ of the research population was at the heart of my research methodology. By that virtue, even the present situation of crossing a flooded area should theoretically have been something I would have had to prepare for. However, faced with the disaster first-hand, I was anything but prepared to encounter the ‘lived experiences’ of my research population. I found myself debating whether loyalty to my research methodology was more crucial than my personal safety or, more importantly, whether being an empathetic researcher and registering the real difficulties faced by the research population, in hosting me in their flooded homes, was the most important objective.

The first thing that was stopping me from stepping into the waters was not the fear of the water itself. We were witnessing people crossing the submerged road. And in all fairness, it was not an ‘alarming flood situation’ by any measure, while perhaps only moving towards that. My concern was one of return: my rapport with these families had not matured adequately to a point that I could stay unannounced in my interviewees’ homes. At the same time, if the rains continued (which was most likely), it would have been risky to return. The families were already struggling with minimal living spaces. Basic amenities like food, drinking water, public transport access, markets, hospitals, etc. were already limited and at far-off distances. With rising waters, it would be inappropriate to obligate them to accommodate an additional person, that person being myself.

I was tiptoeing ahead absent-mindedly with all these thoughts in my head when my auto-rickshaw driver called out, asking me to return to the vehicle. Along with a few passengers, he was planning to return to the main highway. He insisted that I must, too, as I looked like an ‘outsider’ and I wouldn’t be able to cross the road like the ‘locals’. With a sense of self-betrayal, I shut my umbrella and got back to the vehicle.

The next morning, we were informed that the entire road till the bank of the river (located at least 15 kms from where we had returned) had been submerged. Even steamer services to Majuli were shut down indefinitely. I realised my return to meet the families would have to wait. This reflected my limited role as an ethnographic researcher – to study the research population during disasters that very much defined the everydayness of their lives.

The ethnographic project is in itself embedded within power relations between the researcher and the researched (Behar, 1993: 31 cited in Prasad, 1998). Empirically speaking, the research population and I share our homeland (of Assam), culture and language, both literally and figuratively, to a considerable extent. Yet we are anything but parallel in the legitimacies of our respective lives. To begin with, for instance, my family or I have never encountered a disaster first-hand. Concurrently, in my research, it is I who determine the design, selection of site, population for study and methodology. This essentially puts me in a position of power and privilege over the research population I study who, in contrast, had no choice in choosing my research through which to share their everyday lives.

Given this inequality, the power (of the researcher) could end up being wielded against the best interests of the researched in ethnographic studies during or after disasters. The delicate balance between prioritising the research methodology and prioritising the research population then becomes crucial for us as disaster researchers. The power divide and our mandate to negotiate these nuances becomes much more apparent during our fieldwork. Critical reflection (Foley and Valenzuela 2005) at this juncture might prove to be a useful exercise.

My fieldwork experience has underlined that remaining empathetic and putting the interests of the research population facing disasters before our own research methodology is fundamental. The classic ethnographic training for young researchers is to become ‘one among them’ (the research population), given all other factors are in place. Thus, from the start, there silently remains a distinction between ‘them and us’. However, coming from the wider socio-cultural horizon of the researched, local researchers like myself must be trusting of one’s own understanding of issues and instincts for making decisions in the field, even if such decisions do not necessarily fit within the our methodological approaches that have been argued to be rooted in western thoughts. Engaging in other aspects of fieldwork then, for instance making contacts with local experts, especially with researchers based and working in the field site for sustained periods, could be fruitful.

Growing together with our research: prioritising researchers’ self-care

More than a year has passed since the field experience elaborated above. Since returning to safety that day, I keep wondering if the decision was methodologically ethical. My choice that morning reflects the power imbalance between the researcher and the researched very clearly. I call it a power imbalance because I had the choice to not move ahead to meet the research population, whereas they themselves had no choice to leave their flooded home and return on a sunnier, drier day. That is their life. These families continue living in similar conditions of high risk and vulnerability, even today.

Auto Rickshaw in monsoon in Assam
Figure 3: Our auto-rickshaw returning towards the highway (photograph taken by the author, July 10th, 2019).

I made my choice balancing the palpable risks of entering a flooded area and as an ostensibly empathetic researcher. That being said, it was also because I prioritised the safety of the self. Many of our decisions as disaster researchers get shaped by our relationship of accountability to our host organisations (if any) and towards our own families and loved ones. Ensuring our own safety is one such challenging decision.

From the dilemmas of my ethnographic fieldwork, I learnt to appreciate that our research is as much about us as human beings/researchers as they are about understanding research populations. As I examine their lives and they examine mine, we grow together. After all, ours is a social and not a controlled laboratory situation. What I seek to reiterate here is this: many aspects of fieldwork are beyond our control. What is in our capacity, however, is to take care of our own research, the research population we engage with and our own selves.

By self-care, I refer to not just the physical and mental/emotional health safety. Beyond such strictly defined medical aspects of health, I emphasise being self-empathetic throughout the period of research while referring to a researcher’s self-care. This is especially true if we engage with disaster-affected populations over long periods. Having and practising contingency plans for safety prior to the fieldwork and regular communication with supervisory teams and our support system is a must for disaster ethnographic researchers. That being said, a researcher’s self-care must be held dearly by none other than the researcher herself.

Traversing the ethnographic road to meet disaster-affected populations

Upon my return from the flooded area, I found that colleagues at my host organisation had been worried about me, as had been my friends and family. I am glad I retraced my steps, albeit guiltily. In hindsight, I question whether I should have changed my methodology considerably for smoother sailing or perhaps should have conducted fieldwork in areas with minimal probability for a disaster. In that case, how true would I remain to my research question and how ethical would that be? My original fieldnote from that day reads,

… this sight (of the flooded paddy fields all around me) is making me question whether my original study population is even living in the same place where I’d met them, or have they already had to move to avoid the rising river? What, then, does it reflect about the (mobile) lives that these (displaced?) families lead and about the credibility of the gaze I want/need to have for this research?

(‘Reflections’ – Fieldnotes of a missed interview, July 10th, 2019)

Such reflexivity helped me reshape aspects of my fieldwork’s methodology in tune with the dynamic external environment. Physical safety and mental health of aid workers are integral to the everyday conversations of the world of humanitarian aid. While discussing reflexivity in her auto-ethnography with disaster-affected communities of Aceh, Indonesia, Rosaria Indah (2018) shares that secondary traumatic stress (STS) could be one of the long-lasting impacts on disaster ethnographers. Thus, this conversation deserves to pick pace within the academic community too, especially for researchers engaging in long-term humanitarian contexts.

Taking an ethnographic road to disaster-affected communities makes us grow deeply aware of seething worldly inequalities that disasters bring forth. At the same time, it makes us compassionate towards the world outside. It is imperative we reserve a piece of that compassion for our own selves too.

This is an edited version of the article that was originally published on the LSE blog.


Foley, S. and Valenzuela, A. (2005) Critical ethnography: the politics of collaboration. In: N. Denzin and Y. Lincoln, eds. The Sage handbook of qualitative research. 3rd edn. Thousand Oaks: Sage Publications, pp. 217–234.

Indah, R. (2018) Probing problems: Dilemmas of conducting an ethnographic study in a disaster-affected area. International journal of disaster risk reduction 31, pp.799-805.

Kikon, D. (2019) On methodology: research and fieldwork in Northeast India. The Highlander 1(1), pp. 37–40

Prasad, P. (1998) When the Ethnographic Subject Speaks Back: Reviewing Ruth Behar’s. Translated Woman. Journal of Management Inquiry 7(1), pp. 31-36.

Mausumi ChetiaAbout the author:

Mausumi Chetia is a PhD researcher with the ISS. Prior to joining academics, she was working as a development and humanitarian aid professional in India.

Are you looking for more content about Global Development and Social Justice? Subscribe to Bliss, the official blog of the International Institute of Social Studies, and stay updated about interesting topics our researchers are working on.

Print Friendly, PDF & Email
Germany is a deeply racist country―stop pretending otherwise

Germany is a deeply racist country―stop pretending otherwise

While Germany has been lauded for agreeing to take in 1,700 refugees from refugee camp Moria that recently burned to the ground, the country has been cited as a role ...

Covid-19 | Strengthening alliances in a post-Covid world: green recovery as a new opportunity for EU-China climate cooperation?

Covid-19 | Strengthening alliances in a post-Covid world: green recovery as a new opportunity for EU-China climate cooperation?

As nations turn their attention to fighting the economic crisis resulting from the Covid-19 pandemic, green recovery seems to be a good—and perhaps for the first time, possible—option. As climate ...

When children have children: Can postponing early motherhood help children survive longer?

In 2010, approximately 34% of young women in developing countries – some 67 million – married before reaching 18 years of age. An additional 14-15 million women will marry as children or adolescents every year in the coming decades. Child marriages lead to pregnancies and childbirths at an early age, which can have negative consequences for the health of both mother and child. Does the age at which motherhood takes place matter, and can postponing motherhood into adulthood help increase the chances of children surviving beyond five years of age? My study of teen pregnancies amongst Bangladeshi girls shows that age does matter, and it matters quite a lot.

Baby feet and mother's hand

Globally, in developing countries excluding China, one in three girls will probably be married before they are 18, according to UNFPA figures from 2012. Bangladesh has the highest rate of child marriages in Asia (and the third-highest rate worldwide); two in three women marry as children or adolescents in this country. This exceptionally high rate of child marriages in Bangladesh persists despite a minimum legal marriage age of 18 years for women, and it leads to more teenage pregnancies and a care burden for young women. A survey interviewing 72,662 Bangladeshi mothers in 2001 showed that 10% of interviewed women had their first child between the ages of 10 and 14 years, and another 69% of them had their first child between the ages of 15 and 19 (Figure 1).

Figure 1: Age at which Bangladeshi women have their first child

Bangladeshi women age of first child
Figure 1: In total, 72,662 Bangladeshi women who gave birth in at least two different points in time were surveyed for the Maternal Health Services and Maternal Mortality Survey of 2001. This figure shows which fraction of these women gave birth for the first time at a certain age. Source: Trommlerová (2020).

In general, we know that adolescent childbearing is associated with negative health outcomes for both mother and child. The mother faces an increased risk of premature labour, labour-related complications, and death during delivery (Senderowitz, 1995). She  may also suffer an injury, an infection or a serious health limitation, such as an obstetric fistula or perineal laceration (UNICEF, 2001), due to giving birth. These problems are caused both by physical immaturity and poor socio-economic conditions of young mothers, including a lack of access to sufficient antenatal and obstetric care (WHO, 1999).

The child of an adolescent mother is at a higher risk of a low birth weight (Restrepo-Méndez et al., 2015), which is mainly associated with poor maternal nutrition of adolescents during pregnancy (UNICEF, 2001). Low birth weight is, in turn, a frequent cause of death in the first year of life of infants (McCormick, 1985, Sohely et al., 2001). Apart from biomedical reasons, there are additional channels that link adolescent pregnancies to higher mortality in early childhood. These include insufficient access to maternal health care services and lack of experience in taking care of children.

A better understanding of the link between adolescent childbearing and young children’s survival chances is important as scientific evidence can drive policy changes, particularly in enforcing the minimum legal marriage age in Bangladesh. It can also inform the advocacy of changes in cultural practices. Thus, the central question is: Can postponing motherhood of teenage girls help their children to survive beyond infancy or childhood?

Building on previous knowledge, I looked more closely at the impact of adolescent childbearing on the mortality of young children at different ages between 0 and 5 years. The goal was to separate the effect of having a child at an early age from the fact that poorer (and, frequently, less healthy) mothers tend to marry younger and might therefore have less healthy children. The idea is the following: if children born to young mothers suffer higher mortality in early childhood due to biological factors, such as physical immaturity of mothers and the resulting low birth weight of their children, then we should observe different mortality rates not only between children born to adolescent and adult mothers, but also among siblings born to the same mother in her adolescence and adulthood—that is, in different phases of her life.

It turns out that children born to young mothers (child brides in Bangladesh) are more likely to die in the first year of life than their siblings born later on. This is true irrespective of how rich the household is (left graph in Figure 2). Only in poor households, these negative effects extend up to the child’s fifth birthday (black and blue lines in right graph of Figure 2).

In the two graphs below, we see how much lower the probability of death for an infant or a child is if the mother is older than 10-14 years. The age of the mother is displayed in five-year age groups on the horizontal axes. Different lines indicate different income groups (poorest Quintile 1 – black; Quintile 2 – blue; Quintile 3 – green, Quintile 4 – red; richest Quintile 5 – yellow). The percentage of increased risk of early childhood mortality per age group is shown on the vertical axes of the graphs. The left graph depicts infant mortality (up to one year of age) while the right graph shows child mortality (between one and five years of age).

The graphs show that up to one year of age, the income of the family does not really matter (left) while between one and five years of age, a higher income can help outweigh the negative effect of teenage pregnancies (right). The downward trend observable in the left graph is universal for all income groups and it indicates that all children have higher survival chances in the first year of life if their mother is not a teenager. In the right graph, a similar downward trend is observable only for the two poorest income groups, which means that only in poorer families, children of ages one to five have worse survival chances if their mother is a teenager. The three richer income groups show no downward trend (and their slightly upward trend is statistically not important) which means that in richer families, the mother’s age does not really influence her child’s survival if the child managed to survive the first year of life. The graphs are based on my study of adolescent childbearing among Bangladeshi women.

Figure 2: Infant and child mortality effects of maternal age for five different wealth quintiles

Child mortality Bangladesh
Figure 2: Dash-dotted lines mark the average mortality rate to benchmark the effect sizes. Source: Trommlerová (2020).


These results confirm my idea that the effects of adolescent pregnancies on child survival in the first year of life are of biological nature because they are universal. Possibly, they are related to the immaturity of young girls’ bodies and to low birth weight of their children. Beyond infancy, these negative effects remain only in poorer households, which is consistent with the notion that richer households are able to counteract a biologically induced, worse starting position of children born to adolescent mothers by compensatory investments in child health. However, these investments do not become effective until the children reach the age of one year old.

Finally, the estimated effects are substantial in magnitude: for instance, the survival chances of children born to mothers aged 20-24 years are 56% higher in infants’ first year of life and 24% higher when the child is aged between one and five, when compared to their older siblings who were born to young mothers (aged 10-14 years). These effects either persist or become even larger when comparing adolescent maternal age (10-14) to older ages (25-29, 30-34, etc. up to 45-49 years). Importantly, these results remain true also when I exclude older women or first-born children from the sample.

To summarize, I have shown that infants and children have a much better chance of survival when their mothers are adults. The postponement of motherhood into adulthood could help prevent around 12,900 infant and 18,700 under-five deaths annually in Bangladesh, as rough calculations explained in my paper show. These effects can be directly attributed to the practice of child marriages.

This article is based on a recent paper I authored, see here.

Main reference:

S.K. Trommlerová (2020). When Children Have Children: The Effects of Child Marriages and Teenage Pregnancies on Early Childhood Mortality in Bangladesh. Economics & Human Biology 39, 100904.

Other references:

McCormick, M. (1985). The contribution of low birth weight to infant mortality and childhood morbidity. The New England Journal of Medicine 312(2), 82-90.

Restrepo-Méndez, M.C., D.A. Lawlor, B.L. Horta, A. Matijasevich, I.S. Santos, A.M. Menezes, F.C. Barros and C.G. Victora (2015). The association of maternal age with birthweight and gestational age. Paediatric and Perinatal Epidemiology 29, 31-40.

Senderowitz, J. (1995). Adolescent Health: Reassessing the Passage to Adulthood. World Bank Discussion Paper 272. World Bank, Washington DC.

Sohely, Y., D. Osrin, E. Paul and A. Costello (2001). Neonatal mortality of low-birth-weight infants in Bangladesh. Bulletin of the World Health Organization 79(7), 608-614.

UNFPA (2012). Marrying Too Young. End Child Marriage. United Nations Population Fund, New York.

UNICEF (2001). Early Marriage – Child Spouses. Innocenti Digest 7. Innocenti Research Centre, Florence.

WHO (1999). The Risks to Women of Pregnancy and Childbearing in Adolescence. WHO, Division of Family Health, Geneva.

About the author:

Sofia TrommlerováSofia Trommlerová is a postdoctoral researcher in economics at Universitat Pompeu Fabra in Barcelona, Spain. Her main research interests encompass family economics, gender, child health, development economics, and economic demography. In 2017-2018 she was a postdoctoral researcher in development economics at the International Institute of Social Studies (ISS), Erasmus University Rotterdam, The Netherlands.

Are you looking for more content about Global Development and Social Justice? Subscribe to Bliss, the official blog of the International Institute of Social Studies, and stay updated about interesting topics our researchers are working on.

Print Friendly, PDF & Email