Research

Co-authored with Laure de Preux (Imperial College London).
Link to paper (version May 2024): https://dx.doi.org/10.2139/ssrn.4862789  

Amid the current syndemic of obesity and climate change, little is known about the effect of extreme temperatures on dietary behaviour. Using exogenous daily variations in weather and a nationally representative consumer panel in the U.S., we find that extreme heat increases the volume purchased of sugary drinks, with persistent impacts after accounting for inter-temporal purchase shifts. We explore heterogeneous effects and a range of potential drivers, including changes in shopping habits, inter-channel substitutions, retailers' price adjustments, and psychological biases. Results reveal higher impacts among the most vulnerable households and no evidence of long-run adaptation by historical heat exposure. We combine our estimates with output from climate models for 2080-2099. Our projections indicate that climate change may increase sugary drink purchases.

Seminar presentations: Imperial College London (Economics & Public Policy)
Conference presentations: Agricultural and Applied Economics Association (AAEA 2024); Association of Environmental and Resource Economists (AERE 2024); European Association of Environmental and Resource Economists (EAERE 2024); European Health Economics Association (EuHEA 2024); Sustainable Food Systems Symposium (University of Gottingen, 2024); French Economics Association (AFSE 2024); Canadian Economics Association (CEA 2024); Maastricht Workshop on Applied Economics of the Environment (MAEE 2024)

Revise & resubmit at Food Policy. Latest version July 2024.

Solo-authored.

In February 2023, Costa Rica reformed its basic value-added tax basket attracting a reduced rate with the aim to promote more balanced diets. This paper assesses ex-ante the impact of the reform on nutrient availability and household spending across income groups. Price elasticities are estimated using pooled National Household Income and Expenditure Survey data and a two-step censored quadratic almost ideal demand system model accounting for price and total expenditure endogeneity. Nutritional information is derived from food composition tables. Food items are grouped by processing level as a proxy to explore differences in own- and cross-price effects between healthier and less healthy items.  

Seminar presentations: University of Birmingham (Centre for the Economics of Obesity); University of Costa Rica (School of Nutrition); Imperial College London (Centre for Health Economics & Policy Innovation; Public Health Policy Evaluation Unit)
Conference presentations: Quantitative Methods in Social Sciences conference (University of Warwick, 2023)

Work-in-progress. Solo-authored.

Work-in-progress.

Co-authored with Olivier Ecker (IFPRI), Andrew Comstock (IFPRI), and Shaonan Wang (University of Georgia).

Work-in-progress.

Co-authored with Jingmin Zhu (Imperial College London), Jack Olney (Imperial College London), William Joe (IEG, New Delhi), Manika Sharma (RTSL, New Delhi), Lindsay Steele (RTSL, New York), and Franco Sassi (Imperial College London)

The consumption of energy-dense foods and foods high in saturated fats, sodium, and sugars (HFSS) is increasing in India, as well as overweight and obesity. At the same time, a significant portion of the poor still cannot afford a recommended healthy diet. This study aims to assess the effect of fiscal policy scenarios to promote healthier diets on the distribution of household expenditure, nutrient availability, and health outcomes. Own- and cross-price elasticities are estimated using the National Sample Survey Office Household Consumer Expenditure survey 2011-2012 and a quadratic almost ideal demand system model accounting for potential sources of endogeneity. Baseline nutrient availability is derived from Food Composition Tables and purchased quantities are adjusted to December 2022 using the Centre for Monitoring Indian Economy’s Consumer Pyramid data and detailed consumer price index trends. We assess various Goods and Services Tax (GST) scenarios using nutrient profile modelling to differentiate tax rates. We estimate the distributional impact of each scenario on household expenditure and per capita nutrient availability as well as national GST revenue. Finally, we evaluate the impact on diet-related diseases and associated healthcare expenditure using the Health-GPS microsimulation model calibrated with national health survey microdata and estimates from the Institute for Health Metrics and Evaluation. 

Seminar presentations: Institute for Economic Growth (New Delhi, 2023)
Conference presentations: Delivering for Nutrition in South Asia (IFPRI, Kathmandu, Nepal, 2023)