Research

I’m a PhD Candidate at the London School of Economics in the Department of Geography and Environment and the Grantham Research Institute. My research empirically examines how households in sub-Saharan Africa and South Asia are affected by environmental and climate change and studies how they respond using panel surveys and geo-spatial data. My supervisors are Dr Charles Palmer and Dr Sefi Roth. I plan to defend in early 2023.

Here are some things I’ve been working on (happy to send draft papers if you are interested!)

Weather shocks and household impacts: Nigeria’s 2012 floods

Nigeria experienced severe flooding during the rainy season of 2012 which cost $17 billion USD in macro-economic terms. This paper provides a complementary micro-level accounting of the flood using panel survey data across 2,500 agricultural households before and after the flood. I examine multiple definitions of flood exposure (self-reports and satellite imagery) and agricultural outcomes (production and value) to better understand local impacts and their distribution. On average, affected farmers lost around 20% of crop production and 40% of crop value after the flood. But not all households in flooded areas were impacted the same: those who were doubly exposed (live in a flooded area and self-reported a flood) lost up to 75% of crop value. Conversely, households living in flooded areas who were not individually affected reported an increase in crop value around 20% without any change in output. Suggestive evidence is presented that food prices increased in flooded areas, benefitting unaffected farmers selling to local markets. The findings of this paper provide a more nuanced and comprehensive understanding of the local impacts of extreme events and can help inform the design and targeting of post-disaster relief.

Household impacts of Indonesia’s Kerosene-to-LPG Program in West Nusa Tenggara Province

Improving household access to cleaner cookstoves has long been a priority for governments in developing countries. One country which has been a leader is Indonesia, which in 2007 embarked on an LPG Conversion Program to convert 50 million households from using a kerosene or biomass-based stove to using a liquefied petroleum gas (LPG) stove, which is cleaner-burning and more fuel-efficient for cooking. In a quasi-experimental evaluation of the program, I exploit the differential timing of implementation to assess the short-term impacts in terms of health outcomes and spending patterns for households. I focus on one province – West Nusa Tenggara – where one district (Lombok) was treated before the other (Sumbawa). Using a panel household survey and a differences-in-differences and instrumental variables empirical design, I find over 60% of treated households report using the new LPG stove 4 years after the program, which resulted in fuel savings of around 30% for adopters. I do not find robust evidence for improvements in health across a range of outcomes related to indoor air pollution. The findings of this paper suggest that it may be difficult to deliver multiple benefits beyond fuel savings to households in the short-term, and have implications for the design of similar fuel conversion programs currently being planned in a number of developing countries.

Exposure to floods, climate change, and poverty in Vietnam
(with Andrew Smith and Ted Veldkamp)

With 70% of its population living in coastal areas and low-lying deltas, Vietnam is highly exposed to riverine and coastal flooding. This paper conducts a “stress-test” and examines the exposure of the population and poor people in particular to current and future flooding in Vietnam and specifically in Ho Chi Minh City. We develop new high-resolution flood hazard maps at 90 m horizontal resolution, and combine this with spatially-explicit socioeconomic data on poverty at the country and city level, two datasets often kept separate. The national-level analysis finds that a third of today’s population is already exposed to a flood, which occurs once every 25 years, assuming no protection. For the same return period flood under current socioeconomic conditions, climate change may increase the number exposed to 38 to 46% of the population (an increase of 13–27% above current exposure), depending on the severity of sea level rise. While poor districts are not found to be more exposed to floods at the national level, the city-level analysis of Ho Chi Minh City provides evidence that slum areas are more exposed than other urban areas. The results of this paper provide an estimate of the potential exposure under climate change, including for poor people, and can provide input on where to locate future investments in flood risk management.

The multifaceted relationship between environmental risks and poverty: new insights from Vietnam (with Ulf Narloch)

Despite complex interlinkages, insights into the multifaceted relationship between environmental risks and poverty can be gained through an analysis of different risks across space, time and scale within a single context using consistent methods. Combining geo-spatial data on eight environmental risks and household survey data from 2010–2014 for the case study of Vietnam, this paper shows: (i) at the district level, the incidence of poverty is higher in high risk areas, (ii) at the household level, poorer households face higher environmental risks, (iii) for some risks the relationship with household-level consumption varies between rural and urban areas, and (iv) environmental risks explain consumption differences between households, but less so changes over time. While altogether these analyses cannot establish a causal relationship between environmental risks and poverty, they do indicate that Vietnam’s poor are disproportionally exposed. Given growing pressures due to climate change, addressing such risks should be a focus of poverty reduction efforts.

Land and poverty: the role of soil fertility and vegetation quality in poverty reduction (with Martin Heger and Gregor Zens)

The debate on the land–poverty nexus is inconclusive, with past research unable to identify the causal dynamics. We use a unique global panel dataset that links survey and census derived poverty data with measures of land ecosystems at the subnational level. Rainfall is used to overcome the endogeneity in the land–poverty relationship in an instrumental variable approach. This is the first global study using quasi-experimental methods to uncover the degree to which land improvements matter for poverty reduction. We draw three main conclusions. First, land improvements are important for poverty reduction in rural areas and particularly so for Sub-Saharan Africa. Second, land improvements are pro-poor: poorer areas see larger poverty alleviation effects due to improvements in land. Finally, irrigation plays a major role in breaking the link between bad weather and negative impacts on the poor through reduced vegetation growth and soil fertility.