Kathryn Baragwanath

We survey research on the relationship between democracy and the environment. The first part of our review examines how democratic systems influence environmental outcomes. We show scholars find, at best, a weakly positive correlation between democracy and environment, with little support for the proposed Democratic Environmental Kuznets Curve, a finding we confirm with new data. We argue that democracy is too coarse a category to capture variation in environmental outcomes. Therefore, the second part of the review surveys how specific institutional features structure principal-agent relationships between citizens, leaders, and organized groups. We showcase that effective environmental governance depends on institutions that align incentives, reduce informational asymmetries, and match temporal horizons. These can arise even in autocratic contexts, as evidenced by the case of China. We conclude by identifying key open questions and promising directions for future research.

We use a general spatial model to estimate the impact of transportation infrastructure on the spatial distribution of economic activity. We do so limiting ourselves solely to publicly available remotely-sensed satellite data and digitized road maps. Our setting is the expansion of India’s road and highway network during the 2000s. By demonstrating that our approach works in India, where we can validate our results using data from the country’s economic census, we aim to create a framework that is applicable to the many environments where data are even more sparse and (or) where transport networks span national borders.

Satellite data is essential for enforcing and evaluating environmental policy, but technological limitations of monitoring systems can create perverse incentives and bias impact assessment. This study examines how detection thresholds in satellite monitoring systems affect both the implementation and evaluation of forest conservation policies. We identify three key mechanisms: a measurement issue, where datasets with larger minimum detection thresholds systematically miss small-scale deforestation; a loophole effect, where policy only reduces detectable, large-scale deforestation; and strategic adaptation, where regulated agents adjust behavior to exploit known detection thresholds, substituting from large- to small-scale deforestation. Studying Brazil's 2008 municipal Blacklisting policy, we find that the government’s primary monitoring system, which does not report patches below 6.25 hectares, overestimates policy effectiveness by a third compared to datasets with smaller minimum detection thresholds. When measured with those datasets, blacklisting reduced deforestation by 31.2% from baseline-substantially less than the 47.6% reduction suggested by government data. Average clearing size declined by 28.9%, with significant increases in patches below detection thresholds, reflecting both undetected and strategically fragmented activity. Our analysis reveals a critical challenge for environmental governance: as monitoring systems improve, so too do evasion strategies, requiring close attention to how technology shapes observed outcomes and on-the-ground incentives.

Forests serve a crucial role in our fight against climate change. Secondary forests in the form of forest restoration and reforestation provide important potential for conservation of biodiversity and climate change mitigation. In this paper, we explore whether collective property rights in the form of Indigenous Territories (ITs) lead to higher rates of secondary forest growth on previously deforested areas. We exploit the timing of granting of property rights as well as the geographic boundaries of ITs and two different methods, regression discontinuity design and difference-in-difference to recover causal estimates. We find strong evidence that homologated Indigenous territories not only reduce deforestation inside their lands, but also lead to higher secondary forest growth on previously deforested areas. After homologation, ITs displayed higher secondary forest growth than land outside ITs with an estimated effect of 2\%. Furthermore, the average age of secondary forests was 2.6 years older inside homologated ITs. Together, these findings provide evidence for the role that collective property rights can play in the push to restore forest ecosystems.

In this paper, we draw on common-pool resource theory to argue that indigenous territories, when granted full property rights, will be effective at curbing deforestation. Using a novel dataset, we test the effect of property rights on deforestation between 1982 and 2016. In order to identify causal effects, we combine a regression discontinuity design with the orthogonal timing of homologation. We find that observations inside territories with full property rights show a 66% decrease in deforestation, while the effect does not exist in territories without full property rights. While these are local average treatment effects, our results suggest that not only do indigenous territories serve a human-rights role, but they are a cost-effective way for governments to preserve their forested areas. First, obtaining full property rights is crucial to recognize indigenous peoples original right to land and protect their territories from illegal deforestation. Second, when implemented, indigenous property rights create sustainable areas in the Amazon rainforest, providing an important positive externality for Brazil and the rest of the world in terms of climate change mitigation.

This paper proposes a methodology for defining urban markets based on economic activity detected by satellite imagery. We use nighttime lights data, whose use in economics is increasingly common, to define urban markets based on contiguous pixels that have a minimum threshold of light intensity. The coarseness of the nightlight data and the blooming effect of lights, however, create markets whose boundaries are too expansive and too smooth relative to the visual inspection of actual cities. We compare nightlight-based markets to those formed using high-resolution daytime satellite imagery, whose use in economics is less common, to detect the presence of built-up landcover. We identify an order of magnitude more markets with daytime imagery; these markets are realistically jagged in shape and reveal much more within and across-market variation in the density of economic activity. The size of landcover-based markets displays a sharp sensitivity to the proximity of paved roads that is not present in the case of nightlight-based markets. Our results suggest that daytime satellite imagery is a promising source of data for economists to study the spatial extent and distribution of economic activity.

The Green Municipalities Program (Programa Municípios Verdes–PMV) has operated in the Brazilian state of Pará since 2010. Though voluntary, the program has provided \$20M in funding to local governments to combat deforestation in the Brazilian Amazon through improvements to local systems and bureaucracies. By leveraging advanced causal inference methods and twenty years of municipality and property-level land use data, we examine the program’s effects on deforestation, secondary forest regeneration, and local governance outcomes. We find that PMV is associated with a 12\% annual reduction in deforestation and a 7\% increase in secondary forest cover, though deforestation results are sensitive to the inclusion of municipalities previously subject to federal blacklisting. Secondary forest recovery appears more robust and concentrated in younger forests and medium-sized properties, suggesting improved land use incentives and enforcement capacity. We also find evidence of reduced land conflict and increased non-agricultural economic activity. Under conservative carbon valuation assumptions, estimated climate benefits exceed program costs. These findings point to the potential of subnational governance investments to support conservation and sustainable development objectives in tropical forest regions.

This paper examines how domestic economic conditions shape international relations. We develop a novel measure of inter-government interactions using high-frequency event data across 18,330 country dyads from 2001-2019. To establish causality, we exploit plausibly exogenous variation in countries' natural resource wealth driven by global commodity price shocks. We find that positive resource shocks significantly increase countries' aggressive behavior in international relations, primarily through verbal rather than material confrontation. This effect operates strategically: aggression is targeted at peripheral nations while avoiding major trading partners, suggesting a deliberate approach to status enhancement that preserves economic relationships. The mechanism works through domestic political channels, with resource windfalls reducing public discontent and providing governments with political capital to pursue more assertive foreign policy. Consistent with theories of status-seeking behavior as a tool for enhancing international standing, the effects are concentrated in middle and low-income countries and in political systems with electoral accountability. Our findings highlight how domestic economic conditions influence international relations through the strategic pursuit of status, with implications for understanding the economic roots of geopolitical behavior.

Oil royalties provide a substantial and volatile inflow of non tax-payer money to municipal coffers, creating dynamic incentives for politicians in office. Resource windfalls change politicians' budget constraints, generate difficulties for voters to distinguish politicians' integrity, and create incentives for corruptible candidates to enter politics, changing the pool of candidates. Using a formal model with moral hazard and adverse selection, I show how resource windfalls generate the strategic entry of worse candidates into politics, which creates cycles in corruption and reelection patterns. In Brazil, where offshore royalties are determined and allocated exogenously, oil inflows create strong opportunities for corruption. I find strong effects of oil windfalls on corruption. On average, a one standard deviation increase in oil royalties produces a 29% increase in corruption. The effects of windfalls on corruption are larger after elections during booms and lower during busts. Furthermore, oil royalties lead to a reelection cycle: when the price of oil is expected to be higher, incumbents are reelected more often than when the price of oil is expected to fall, independent of economic and individual level variables. I show that strategic entry of worse candidates during booms is likely the cause of these corruption and reelection cycles, as predicted by the theory. Taken together, these results point to a strong effect of oil royalties on local level political equilibria.

Women have been shown to have different policy preferences, invest in different types of goods and be less corrupt than men when elected into political office. In this paper, we study the effects of electing a woman mayor into office in the Brazilian Amazon on rates of deforestation. We exploit close election regression discontinuity design in order to establish causal findings. We find that electing a woman mayor leads to significantly lower deforestation rates during the women's time in office. We propose that women's effect on reduced deforestation comes through two mechanisms: (i) women's distinct preferences towards climate change and (ii) a lower likelihood of being captured or corrupt. Although we can't fully differentiate between the mechanisms, we provide evidence that the corruption mechanism is certainly at play. Women mayors are less corrupt than their male counterparts, receive less campaign funding from businesses, and are less likely to receive funding from large agricultural and mining companies. Altogether, our findings strongly support the theory that women's representation reduces deforestation rates in the Brazilian Amazon, with evidence that lower likelihood of regulatory capture and corruption levels are driving these results, at least in part.