Research
My research uses experimental, econometric, and machine-learning methods to estimate the causal effects of policy on mobility, sustainability, and welfare. Where available, abstracts and PDFs are linked below.
Working papers
Income Shocks and Intra-household Bargaining: Theory and Evidence from South Africa
Submitted · 2025
Abstract
I extend the revealed-preference framework of Cherchye et al. (2015) for the household sharing rule — the share of resources allocated to each decision maker — from two members to m. This matters because households of three or more co-resident adults are common in lower-income settings, and the two-person version mechanically overstates an individual’s share. I apply the extended framework to the South African Older Persons Grant, a large means-tested transfer that acts as a plausibly exogenous income shock, using a fuzzy regression discontinuity design on five waves of panel data. Pension receipt raises the recipient’s sharing-rule bound by about 8% (income elasticity ≈ 0.15), with gains concentrated among female and unmarried recipients. Receipt also lowers labour-force participation for the recipient and other household members, leaving total household income roughly unchanged.Work in progress
These are preliminary, incomplete drafts. Please do not cite or circulate without permission.
Carbon Pricing, Sectoral Channels, and Distributional Incidence: Evidence from Germany’s BEHG
Abstract
Germany’s 2021 carbon price on transport and heating fuels (the BEHG) preceded the EU-wide ETS2, providing early evidence on how such a price affects emissions and households. Using a synthetic difference-in-differences design across Germany’s 401 districts, with European regions as a donor pool, I decompose the policy’s effect into a transport and a residential heating channel. The estimates indicate a robust reduction in transport emissions, while the residential channel displays a counterintuitive increase in particulate concentrations consistent with a shift toward biomass heating during the gas crisis. The paper further quantifies consumer pass-through, the distributional incidence of the burden across districts, and the role of Klimageld revenue recycling. Austria’s analogous carbon-pricing scheme serves as an independent comparison case.Crisis-Born Capital: Endogenous Investment and the Private Response to South Africa’s Loadshedding Crisis
Abstract
This paper examines how private capital responds when public infrastructure fails, using South Africa’s 2019–2024 electricity load-shedding crisis. I construct an area-level panel of outage hours from the revision history of a crowdsourced load-shedding schedule and combine it with customs data on solar and battery imports, national electricity statistics, and a satellite-based inventory of installed solar capacity. I estimate the response of private backup investment to outage exposure, separating a reliability channel from the concurrent decline in global photovoltaic prices. Preliminary results indicate that both outages and falling prices drive adoption, pointing to a distinct reliability motive. A cross-country comparison situates South Africa among the most severe episodes of crisis-induced private capital formation internationally.Colonial Cash-Crop Lock-In as a Climate Liability: Threshold-Crossings in Former Plantation Economies
Abstract
This paper examines whether colonial-era agricultural specialisation has become a source of climate vulnerability. Where colonial administrations entrenched heat-sensitive cash crops, regions may face disproportionate losses as warming pushes those crops beyond their biological thresholds. Combining a digitised colonial cash-crop map of Africa with satellite night-lights, historical temperature records, and climate projections, I estimate within-country relationships between colonial cash-crop status and realised warming. The estimated effects concentrate among the crops and regions for which the threshold mechanism is most plausible, and a comparable pattern emerges for rubber cultivation in Southeast Asia. Work is ongoing to sharpen identification and to translate the estimates into welfare terms.The Imported Grid: Long-Run Effects of Colonial Urban Planning Doctrine on Within-City Flood Vulnerability
Abstract
This paper asks whether colonial urban-planning doctrine has had lasting consequences for flood risk. British colonial planners frequently imposed a standardised, London-derived street grid on cities irrespective of local topography. Using repeated satellite flood imagery and high-resolution elevation data, I compare flooding immediately inside and outside the planned colonial core across several dozen cities, and test whether the difference increases with the mismatch between the imposed layout and local terrain. The boundary effect is robust across alternative boundary definitions, including legally defined cantonment limits, and a climate-based placebo is null. Current work focuses on inference and on quantifying the associated welfare costs.Policy-Restricted Residential Sorting and Urban Amenity Capitalization: Evidence from the South African Group Areas Act
Abstract
This paper uses South Africa’s Group Areas Act, which restricted residential location by race from 1950 until its repeal, to identify preferences for urban amenities and the welfare costs of policy-restricted sorting. I reconstruct the statutory boundaries for Cape Town by text-mining historical Government Gazettes, yielding sharp discontinuities in racial composition and local amenities, including air quality. The empirical strategy combines a boundary regression-discontinuity design with a discrete-choice residential sorting model, complemented by a panel analysis tracking the evolution of racial income gaps following repeal. The project is in progress, with current work centred on the sorting model.The Revealed Local Value of Big-Game Wildlife: Evidence from 10,000 Protected Areas
Abstract
This paper estimates the revealed local economic value of big-game wildlife across roughly ten thousand protected areas, drawing on several independent identification strategies rather than a single estimate. The first compares the spatial clustering of tourism activity around big-game reserves with that around comparable non-tourism reserves; the second exploits the 2008–2014 poaching wave to test whether built-up development near reserve boundaries tracks wildlife abundance; the third examines the cropland- and forest-free buffers that reserves sustain. These moments are combined with a spatial model of tourism-operator location choice. Preliminary estimates suggest a local value substantially exceeding typical protected-area operating budgets, though this remains subject to further testing.Publications
Pigovian Transport Pricing in Practice
Review of Economic Studies (forthcoming, 2026) · CESifo Working Paper No. 11871
Abstract
We implement Pigovian transport pricing in a field experiment in urban agglomerations of Switzerland over the course of 8 weeks. The pricing considers external costs from climate damages, health outcomes and congestion and varies across time, space and mode of transport. The treatment reduces the external costs of transport of the treated individuals by 4.6% in the short run. The main underlying mechanism is a shift away from driving towards other modes, such as public transport, walking and cycling. Providing information about external costs alone changes behavior of altruists, but not for the whole sample. The welfare improvement due to mode shift is around 77 Swiss francs per person and year, of which a revenue-equivalent fuel tax would achieve about 70%.The Impact of COVID-19 on Mobility Choices in Switzerland
Transportation Research Part A: Policy and Practice · 2023
Abstract
We study the effect of the COVID-19 pandemic and the associated government measures on individual mobility choices in Switzerland. Our data is based on over 1,600 people for which we observe all trips during eight weeks before the pandemic and until May 2021. We find an overall reduction of travel distances by 60 percent, followed by a gradual recovery during the subsequent re-opening of the economy. Whereas driving distances have almost completely recovered, public transport remains under-used. The introduction of a requirement to wear a mask in public transport had no measurable impact on ridership. The individual travel response to the pandemic varies along socio-economic dimensions such as education and household size, with mobility tool ownership, and with personal values and lifestyles. We find no evidence for a significant substitution of leisure travel to compensate for the reduction in work-related travel.The MOBIS Dataset: a Large GPS Dataset of Mobility Behaviour in Switzerland
Transportation · 2022
Abstract
This article presents the MOBIS dataset and underlying survey methods used in its collection. The MOBIS study was a nation-wide randomised controlled trial (RCT) of transport pricing in Switzerland, utilising a combination of postal recruitment, online surveys, and GPS tracking. 21,571 persons completed the first online survey, and 3680 persons completed 8 weeks of GPS tracking. Many continued tracking for over a year after the study was completed. In the field experiment, participants participated through the use of a GPS tracking app, Catch-my-Day, which logged their daily travel on different transport modes and imputed the trip segments and modes. The experiment lasted 8 weeks, bookended by two online surveys. After the first 4-week control phase, participants were split into two different treatment groups and a continued control group. An analysis of the survey participation shows that the technology is capable of supporting such an experiment on both Android and iOS, the two main mobile platforms. Significant differences in the engagement and attrition were observed between iOS and Android participants over the 8-week period. Finally, the attrition rate did not vary between treatment groups. This paper also reports on the wealth of data that are being made available for further research, which includes over 3 million trip stages and activities, labelled with transport mode and purpose, respectively.Observed Impacts of the COVID-19 First Wave on Travel Behaviour in Switzerland Based on a Large GPS Panel
Transport Policy · 2021