Research

A long-run evaluation of the protected areas network in Brazil's mangrove forests 

Abstract: Protected areas (PAs) are a critical tool for meeting conservation and climate change goals, yet significant challenges remain in estimating their true impact on ecological outcomes. Namely, siting considerations affect measured PA effects, highlighting the need for quasi-experimental impact evaluation methods. Further, effects of protection (e.g,, avoided deforestation and regrowth) often only appear many years following PA creation, and vary significantly across ecosystems and temporally. To address these questions and provide a long-run estimate of PA impacts in a habitat of major conservation interest, I construct a dataset of protected area creation and land cover use across 1km grid cells spanning over thirty years along the coast of Brazil, focusing on a common governance type, strongly protected PAs, and ecosystem, mangrove forests. To precisely estimate impacts over time after PA implementation, I use recent innovations in difference-in-difference methodology with staggered implementation (Gardner 2021), enabled by the high temporal resolution of yearly mangrove land cover data (Diniz 2019, Mapbiomas 2024). I find consistently positive effects, ranging from a 1-3 percentage point increase in the likelihood of forest cover after protection. However, the event-time estimates indicate that anticipation may be present, with a significant decrease in forest cover 10-15 prior to PA creation, followed by a recovery to a net increase in forest cover approximately 10 years prior. Differing patterns of effects are also observed across two PA categories. Federal PAs appear to produce an initially stronger effect of protection than state and locally managed PAs, as do PAs which prohibit recreation uses. However, effects appear to dissipate approximately 10 years following PA creation in most categories, with some variation. The results provide strong evidence for positive PA impacts on mangrove forests, adding to the literature on the topic in a specific ecosystem over a wide geographic scale and long time period. At the same time, the findings highlight the presence of a time-specific pattern of anticipatory exploitation and recovery, as well as variability across protected area types and rules.  The measured positive impact underscores the importance of PAs as a policy tool, and for further research which explores the heterogeneities in effects observed.

Economic planning for restoration investments under climate uncertainty: application to the Atlantic Forest of Brazil

Abstract: With the UN declaring the 2020s the “Decade on Ecological Restoration,” interest and funding for restoration projects are steadily increasing (Jones 2018). While the body of ecological research is sizable, the decision of where and when to invest in restoration is fundamentally an economic one. Recent large scale analyses have focused on the prioritization question from a static cost-benefit perspective (Strassburg 2020), an important first approximation . I develop an  economic model of two main extensions of the restoration planning problem: projections of changing conditions, and uncertainty in those projections, for both the investment decision at a single site and prioritization across multiple sites. I first present a generalized stochastic-dynamic framework that can be applied to various contexts. I next apply the model to a restoration practice with a developed empirical literature, Atlantic Forest plantations in Brazil. I focus on two main projections, in tandem with ecological data and theory on regrowth trajectories: economic changes in agricultural opportunity costs and shifts in forest cover, both driven by climate change. Finally, I use estimations of uncertainty in these projections to simulate changes in restoration decisions over time, exploring the implications of those changes on potential drivers of restoration success.

Governance Practices in the Marine Protected Areas of the Sunda Banda and Birds’ Head Seascapes, Indonesia

Abstract: The expansion of Marine Protected Areas (MPAs) is a major conservation priority, providing vital services for local populations and ecosystems alike. However, MPA effectiveness varies widely, often depending on governance—the rules and practices that regulate resource use. In Eastern Indonesia, the State University of Papua (UNIPA) and the World Wildlife Fund have been collecting data from 230 settlements across 12 MPAs to investigate patterns of governance and understand the factors behind MPA success. This study applies Ostrom’s Design Principles for common pool resource governance, aggregating survey data to measure alignment with governance principles and conducting robustness checks to validate the framework. We use cluster analysis to identify common governance patterns among settlements. Our analysis reveals substantial variation in governance across and within MPAs. While some MPAs demonstrate uniformly high levels of governance alignment, others show significant heterogeneity, potentially impacting their effectiveness. Cluster analysis identified four general categories of governance, highlighting areas of both strong and weak alignment with principles. Notably, the role of higher-level governance exhibited a distinct distribution, suggesting external management influences. These findings demonstrate the variation within single MPAs, emphasizing the need for context-specific approaches to governance. Our framework provides a template for evaluating governance practices in MPAs, with broader implications for conservation policy.