Research
Working Papers
Governing Environmental Markets: Evidence from Irrigation in Water Markets
Abstract
Water resources present a classic tragedy of the commons that is of increasing relevance due to climate change. This paper provides evidence of how property rights institutions, particularly local irrigators’ organizations, impact water markets’ efficiency. Our analysis is based on a unique dataset that integrates administrative records, hydrological measures, geographic information, and satellite imagery. We develop a novel misallocation test, which suggests that these organizations reduce misallocation caused by the natural capacity of upstream users to over-extract. We show that these efficiency gains are a result of both water redistribution and individual adaptation, as downstream farmers increase their water consumption and agricultural yield. Large farms extend their growing season, adopt more efficient irrigation technologies, and overall gather more benefits from the analyzed property rights institution. Meanwhile, although upstream farmers reduce their water consumption, their productive outcomes remain unchanged. We also document increases in river streamflow during the irrigation season, concentrated in basins with higher agricultural activity. Our results provide micro-evidence of the consequences of effective governance for both allocative efficiency and equity.
Powering Through Drought: How Electricity Systems Respond to Water Scarcity
Abstract
Changing weather patterns pose a fundamental challenge for countries seeking sustainable development while relying on water-dependent electricity systems. Using two decades of plant-level generation and fuel input data matched to hydrological basins in Chile, we show that local droughts have reduced hydropower output by about 25% on average and that the shortfall has been compensated by thermal plants with spare capacity responding chiefly to system-wide scarcity. This substitution has kept supply flowing but has failed to insulate the system: Wholesale prices rise by about 14% per standard deviation of drought, and exposure to distant drought—which can reach a locality only through the grid—raises service interruption minutes by about 20%. The substitution is also carbon-intensive. Emissions from thermal plants with high spare capacity rise by roughly 29% during system-wide droughts relative to those of comparable plants in the same market and month, and because the compensating generation is only partly redistributed across the network, drought-affected regions can face both water scarcity and higher local pollution. Over longer horizons, droughts have been followed by expansions in both thermal and wind capacity, not hydropower—evidence of the market diversifying under stress and, in doing so, deepening carbon lock-in. These results trace how hydrological risk propagates from generation to prices, reliability, and the capacity stock.
Trade and Structural Change: Evidence from Colombia and the Panama Canal
Abstract
This paper leverages a quasi-natural experiment to trace the short- and long-run effects of international trade on structural change in an economy with a comparative advantage in agriculture. I find that regions in Colombia that gained access to international markets thanks to the opening of the Panama Canal experienced higher population growth and faster structural change. While the impact of trade on population growth is increasing in the capacity to grow agricultural exports, the impact of trade on structural change is maximized at intermediate levels. Furthermore, I provide evidence of the indirect effects of trade: regions located between ports and places suitable for growing agricultural exports saw a larger share of their labor reallocated towards services. The findings highlight the within-country distributional consequences of exposure to international trade for developing countries.
Bimodal Transportation Infrastructure and Regional Development: Evidence from Argentina, 1960–1991
Abstract
This paper estimates the impact of alternative transportation modes on economic development. It focuses on Argentina, where over three decades, 10,000 kilometers of railroads were closed, and 18,000 kilometers of paved roads were constructed. Our empirical strategy leverages a discontinuity in how a World Bank mission selected railroads for closure assessment. We find that the dismantling of railroads operated as a negative trade shock that led to an overall decline in economic activity. It decreased the share of labor in agriculture - the export sector - and had negative impacts on both the population and industrial production. We find evidence of paved roads increasing the share of labor in manufactures, but no impacts on population growth nor industrial production.
Selected Work in Progress
Balancing Act: Hydropower Generation and Competing Water Uses
Drought and the Power Mix in Latin America: Short-Run Substitution and Long-Run Investment
Abstract
Hydropower generates two fifths of Latin America’s electricity, while the region’s droughts grow more frequent and more intense. This paper measures their effects on power supply in three steps: the hydropower lost, how the shortfall is covered, and what gets built afterwards. We use a monthly panel of 2,814 plants in ten systems of Brazil, Chile, Colombia, Ecuador and Peru, 70% of the region’s hydropower capacity, for 2000–2024, and measure drought with runoff anomalies in each plant’s basin. One standard deviation of scarcity lowers a system’s hydro output by 8.8% of mean generation; a basin in drought loses a sixth of its hydro output within six months, and its reservoirs lose more than its run-of-river plants. Where a system can trade, imports cover part of the loss. Every system burns more thermal fuel: thermal output rises 19%, at plants burning diesel, fuel oil or a mix of gas and oil, mostly lightly used plants that are already running, and not at gas plants. At fixed emission factors, CO₂ rises 18% and SO₂ 23%. After three years of scarcity, new thermal plants enter in Brazil and Chile; in Colombia, new hydro follows a single dry year. What a system emits in dry years depends on the backup plants it keeps.
Who Turned On the Lights? Electricity Supply Shocks and Development in Ecuador
Abstract
We estimate the causal effect of electricity-service reliability—the quality of an existing connection rather than access to it—on local productive activity in Ecuador, where hydropower supplies close to 80 percent of generation. We measure reliability from the regulator’s interruption records and estimate dynamic responses with a local-projections instrumental-variables design, instrumenting municipal reliability with a shift-share instrument based on river runoff anomalies (deviations from historical means), weighted by upstream hydroelectric capacity and distance. The most robust response is compositional: improved reliability shifts the local firm population toward younger firms—average firm age falls, consistent with rising entry and falling exit—with the clearest activity response among young firms, whose exports rise over the medium run. Local sales, exports, payroll and wages move in the same direction but are individually imprecise, while employment shows no response. Because the identifying variation is drought-driven, the results also speak to how climate variability reaches firms through the electricity channel in hydro-dependent economies.
A Tale of Two Pathways: The Local Economic Impacts of Solar Energy Investments in Colombia
Abstract
Solar investment comes in two forms: large plants connected to the grid and small systems installed by households and firms for their own use. We study their local economic effects in Colombia, where between 2017 and 2024 small plants were 99% of the solar plants installed and large plants 87% of the new capacity. We combine plant-level data on solar investment with a municipal panel of formal employment, tax revenue and value added, and estimate local projections separately for plants above and below 1 MW. Large plants follow the cycle of a construction project: construction employment and tax revenue rise while they are built and employment falls after start-up, value added rises in the primary sector, and neighbouring municipalities lose workers. Small plants raise formal employment one year after start-up, across trade and services, and raise municipal tax revenue, largely through the tax on local business income, with no effect on total measured value added and positive spillovers to neighbouring municipalities. The local footprint of solar deployment depends on the scale at which it is installed.
Agrarian Reform, Structural Transformation and Regional Development in Post-Revolutionary Mexico, 1916–1976