Urbanisation generates powerful agglomeration benefits but also severe congestion externalities that require regulation and public goods provision. In developing megacities, governments lack the fiscal and enforcement capacity to implement textbook solutions. In their absence, unregulated private provision emerges across sectors such as waste, sanitation, water, and transport, leaving externalities largely unaddressed. What can economics offer cities that cannot implement textbook policies? We study whether limited, feasible regulation can leverage informal markets to deliver public services at scale, focusing on trash collection in Accra, Ghana. Using novel data, we document an imperfect but functioning market: households pay informal tricycle collectors for door-to-door collection, yet prices remain too high for universal access, and final disposal occurs largely at illegal dumpsites. We identify a single feasible policy lever—subsidies for formal, environmentally controlled disposal—and evaluate it using a new structural model of urban waste collection and disposal. To estimate the model, we combine survey experiments to elicit household demand, a custom app to record transactions and tricycle routes, and a field experiment randomising dumping fees. Counterfactual results show that a modest subsidy achieves roughly 80% of the efficient outcome, substantially reducing environmental damages while generating larger welfare gains at lower cost than planned infrastructure expansion. These findings demonstrate that even in very low-capacity environments, limited price-based regulation can align informal private initiative with social objectives.

Presentations: (2025) LSE Environment Camp, LSE IO-DEV Workshop, Seventeenth Annual Conference on Urban and Regional Economics (CURE), Guest Lecture at Kyiv School of Economics' Masters program in Urban Studies and Postwar Reconstruction; (scheduled) STEG Annual Conference and Theme Workshops 2026

Funders and Partners: International Growth Centre (IGC), STICERD, Innovations for Poverty Action (IPA), DataPlas Ghana

Do industrial policies that promote clean energy offer a "ray of hope" increasing a country's economic growth and welfare whilst simultaneously reducing carbon emissions? We study the impact of Chinese solar subsidies whose implementation by cities coincided with a dramatic fall in global solar prices. We construct new panel data on city-level solar policies, patenting and output. Using synthetic-difference-in-differences between 2004-2020, we find that production and innovation subsidies were more effective than demand-side (installation) subsidies in generating large and persistent increases in local innovation, firm numbers, output and exports. However, demand policies are most effective at reducing pollution. We build and estimate a quantitative model with endogenous innovation that takes into account business stealing and knowledge spillovers. Through the lens of this model we show that: (i) the local effects remain substantial at the aggregate level; (ii) Industrial policy explain almost two-fifths of the decline in the global cost of solar panels and over a third of the increase in Chinese innovation; (iii) Innovation subsidies are much more cost-effective than demand and production subsidies; and (iv) solar industrial policies increased Chinese welfare by 1% to 2.3%, almost as much as existing estimates of WTO Accession.

Presentations: (2023) NBER Summer Institute EEE, Stanford SITE; (scheduled) ASSA Annual Meeting in Philadelphia

The internal structure of cities is shaped by localised production and residential externalities. Lack of sanitation gives rise to negative externalities in the form of pollution and disease transmission. Sanitation investments might therefore be essential to allow for people to tolerate density and for positive agglomeration externalities to emerge; thereby turning population density from a downside into an upside. To investigate this hypothesis, we assemble an address-level data set spanning the entire construction period of the Parisian sewage network during the 19th century. We first present event studies examining changes in firm density, residential density, land values, and disease incidences. We then leverage an urban model to estimate the impact of sewers on the strength of agglomeration and city structure. Through counterfactual analyses we examine the impact, cost-benefit, and alternative placements of sewers. Finally, we asses the extent to which sewers displaced pollution to the Parisian suburbs, affecting their long-term development.

Economic growth involves the spatial and sectoral reallocation of people and firms. How much should governments plan this reallocation, and what makes spatial planning successful? South Korea rose from poverty to become a leading industrialised economy today in a remarkably condensed timeline. The "Miracle on the Han" involved rapid urbanisation, industrial modernisation, and rural productivity growth, alongside aggressive spatial planning policies in the form of regional industrial parks, highway construction, and rural infrastructure development. We are creating a new historical database tracking all facets of South Korea's structural transformation from the mid- to late-20th century. Utilising newly digitised historical micro data on output, firms, employment, population, urbanisation, and infrastructure, we study the policies and channels through which the country achieved its remarkable transformation.

Funders and Partners: Structural Transformation and Economic Growth (STEG), Economic and Social Research Council (ESRC)

Electricity is a crucial input for modern production, yet its quality across time and space varies substantially. In this project, we ask how the endogenous allocation of electricity affects firm productivity and agglomeration. We build a unique spatial dataset combining administrative data on electricity consumption, engineering records on electricity outages, and a detailed firm panel from a national census. With this data in hand, we use a machine learning algorithm to predict the spatial allocation of outages following exogenous supply shocks. Using this as an instrument, we obtain reduced-form causal estimates of how firms respond to outages. Lastly, leveraging a dynamic spatial general equilibrium model, we compare the current equilibrium observed in the data to counterfactuals where outages are (i) reduced and, (ii) optimally adjusted to maximise welfare, taking into account the spatial misallocation in firm entry and growth across the country.

Funders and Partners: International Growth Centre (IGC)

Using geolocated information on over 400,000 water points and a variety of health surveys across Sub-Saharan Africa, we document that access to functioning water and improved sanitation increases villages' resilience to high temperatures, reducing mortality and disease incidence. Leveraging machine learning techniques and rich information on the characteristics of water points, we identify that whether maintenance is private, communal, or institutional is a key determinant for a water point functional status. We design an intervention to improve the maintenance of water points when its responsibility is assigned at the community-level.

Funders and Partners: International Growth Centre (IGC)