Journal Articles:
The persistence of trade relocation from civil conflict
with Tobias Korn, Journal of Development Economics, 2025.
This paper examines the lasting impact of civil conflicts on bilateral trade flows and the subsequent implications for economic recovery. Utilizing a novel estimation approach based on the structural gravity model of international trade, we demonstrate that importers shift their trade preferences away from exporters involved in civil conflicts. This effect persists even after the conflict has been resolved, as countries solidify their relocation decisions by reducing bilateral trade costs with alternative trading partners through Preferential Trade Agreements. Notably, the persistent trade relocation is more pronounced in the manufacturing sector, while it does not occur in the fuels sector. Our findings underscore the significance of supportive trade policies as effective tools for assisting nations in recovering from episodes of political violence. Furthermore, our estimation approach can be adapted to investigate the impacts of other unilateral shocks, such as natural disasters, or to analyze various bilateral dependent variables, including migration.
Disability and risk preferences: Experimental and survey evidence from Vietnam
with Jan Priebe and Ute Rink, The Economic Journal, 2024.
We investigate whether experiencing a disability incidence in the household affects economic risk preferences in Vietnam, leveraging (i) ten years of individual-level panel data and (ii) data from a lab-in-the-field experiment. We find that individuals who experience a disability event in the household behave in a more risk-averse manner than individuals without such an experience. Examining potential underlying mechanisms, we demonstrate that a household disability shock leads to lower wealth, which in turn is related to higher levels of risk aversion. Furthermore, we provide evidence that cognitive mechanisms—fearful emotions and the updating of beliefs (becoming more pessimistic about the future)—are another, perhaps even more important channel through which disability shocks affect risk preferences.
Better data for decent work in the global food system,
with Eva-Marie Meemken et al., Nature Food, 2024.
Working conditions in food systems remain precarious across the globe. Little scientific guidance exists on what works where when it comes to initiatives aimed at addressing this issue. Investments in large-scale, nationally representative agricultural worker data are needed to properly document the scale and nature of working conditions and better guide policy design and implementation.
World Development, 2023.
There are growing concerns that automation technology may have far-reaching implications for development, by restructuring global value chains and substituting workers. This paper investigates how domestic and foreign automation impact a resource-rich emerging economy. The empirical analysis builds on a Ricardian model of trade and a shift-share approach. Differences in regional industrial compositions are used to translate domestic industry-level robot adoption to local labor markets in Brazil. Differential trade and inter-sectoral input–output linkages between a foreign industry and regions in Brazil are leveraged to construct a measure of exposure to foreign automation. Instrumental variables account for endogeneity in robot adoption. Larger exposure to foreign automation is found to decrease the share of manufacturing employment and increase the share of employment in the mining sector. These shifts are driven by changes in the demand for export goods from local labor markets. Domestic automation benefits higher skilled and female workers. The findings suggest that foreign automation may contribute to “premature deindustrialization” in emerging economies.
Greenhouse Farming and Employment: Evidence from Ecuador,
with Eva-Marie Meemken, Food Policy, 2023.
The transforming agri-food sector is an important contributor to employment generation in rural areas of lower-income countries. However, little attention has been paid to the question of how technology affects the quality and quantity of employment in the sector. In this paper, we provide the first empirical evidence of how the rise in greenhouse farming changes labor demand (i.e., the number of employees), using large-scale and nationally representative agricultural survey data from Ecuador between 2014 and 2021 and pseudo-panel as well as event-study estimation techniques. Contrary to fears that more technologically advanced production methods displace large shares of the workforce, we find that greenhouse farming is associated with higher labor demand. Specifically, greenhouse farms hire more female workers and workers on permanent contracts.
Working Papers:
Stress Testing Survey to Survey Imputation: Understanding When Poverty Predictions Can Fail
with Paul Corral, Andres Ham, Peter Lanjouw, Leonardo Lucchetti, World Bank Policy Research Working Papers, 2025.
Accurate and timely poverty measurement is central to development policy, yet the availability of up-to-date high-quality household survey data remains limited—particularly in countries where poverty is most concentrated. Survey-to-survey imputation has emerged as a practical response to this challenge, allowing practitioners to update poverty estimates using recent surveys that lack direct welfare measures by borrowing information from other comprehensive surveys. A critical review of the method is provided, revisiting its statistical underpinnings and testing its limitations through extensive model-based simulations. Through these simulations, the analysis demonstrates how violations of parameter stability, omitted variable bias, and shifts in survey design can introduce substantial errors—particularly when imputing across time or under economic and structural change. Results show that standard corrections such as re-weighting or covariate standardization may fail to eliminate these biases, especially when imputing across time or under structural change. The performance of alternative model specifications is also evaluated under various methods, including performance under heteroskedastic errors, non-normality. The findings offer practical guidance for practitioners on when survey-to-survey imputation is likely to succeed, when it should be reconsidered, and how to communicate its limitations transparently in the context of poverty monitoring and policy design.
Does Outgrowing equal Outsourcing? Contract Farming and Labor Demand
with Eva-Marie Meemken
Contract farming—a pre-harvest agreement between farmers and buyers such as processors, retailers, or large farms—has received substantial attention in the scientific literature and development policy debate. Yet the extant literature focuses almost exclusively on welfare implications for smallholder farmers participating in these schemes, paying little attention to the other side of the contract. We address this gap by estimating how contracting affects labor demand on large farms, employing an instrumental variable (IV) approach, and using unique census data from Tanzania. The instrument exploits historic variation in local cropping patterns during the time period when contract farming expanded rapidly in Tanzania. The empirical results show that contracting increases the demand for higher-skilled technical workers on large farms, while it makes it more likely to not hire any lower-skilled laborers. Further analyses suggest that the provision of extension services and different motives for engaging in contracting may be mechanisms behind the results. The findings are relevant for policy and generate several interesting questions that could be explored in follow-up research
Dealing with Agricultural Shocks: Income Source Diversification through Solar Panel Home Systems
with Friederike Lenel, Krisztina Kis-Katos and Christoph Weisser
This paper investigates whether small-scale solar panels can help farmers in rural areas of developing countries to mitigate income losses when they experience climatic shocks. We exploit a large dataset containing information about loan repayment as well as electricity usage from solar panels of individuals in Tanzania. Using machine-learning techniques we classify solar panel usage into private consumption and usage for business. We link customer locations to climatic stress events, making use of spatial and temporal variation in climatic shocks. We find that being exposed to climatic shocks reduces the ability of farmers to continue loan repayments. We then explore whether using solar panels for off-farm income generation can mitigate the income shocks.
Policy paper contributions:
World Bank 2023. Digital Progress and Trends Report, World Bank, Washington DC, DOI: 10.1596/978-1-4648-2049-6.
Stemmler, H. 2022. The effects of COVID-19 on businesses: key versus non-key firms, ILO Working Paper 77 (Geneva, ILO): https://www.ilo.org/wcmsp5/groups/public/---dgreports/---inst/documents/publication/wcms_855012.pdf
World Bank 2021. Socioeconomic Impacts of COVID-19 in Kenya on Firms : Rapid Response Phone Survey, Round 1. World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/35172
Work in Progress:
Taking over the World? Automation and Market Power
with Richard Haarburger and Florian Unger
This paper studies how automation technology affects market power in the global economy. We develop a theoretical model in which firms' mark-ups are endogenous to the rate of technological adoption, but are also affected by technological adoption of other domestic and foreign firms. In an empirical analysis, we find that market-power, measured as the mark-up of price over marginal cost, is on average decreasing with a higher rate of automation. There is however substantial heterogeneity, with firms in the highest income quintile gaining market power. On the other hand, larger exposure to foreign automation increases average mark-ups, driven by exports in intermediate goods. These findings suggest adverse effects of automation on competition and market power.