This paper develops a growth model with two social classes—capitalists, who invest in productive assets and housing, and workers, who supply labor and invest only in housing—while distinguishing residential land, residential structures, and productive capital. The model shows that the relative price of land grows with GDP in the long run, while the quantity and price of housing services rise more slowly. The results indicate that shifting taxation toward land property increases long-run GDP growth and produces more egalitarian income and wealth distributions. Allowing a larger share of residential investment to be tax deductible reduces inequality, whereas giving greater utility weight to housing services leads to long-run distributions more favorable to capitalists. Policies or preference shifts that increase investment in residential rather than productive assets generate balanced-growth paths with higher wealth-income ratios. Finally, endogenous fluctuations may arise along the equilibrium path to balanced growth in a model driven solely by fundamentals.

As artificial intelligence accelerates demand for data centers, digital transformation is becoming a land, grid, and governance problem. This paper examines whether countries’ ability to accumulate data-center-related infrastructure depends on land and property governance, infrastructure reliability, and middle-income structural positioning. Using a country-year panel from UN Comtrade, World Development Indicators, Worldwide Governance Indicators, and the World Bank Doing Business historical archive, the analysis distinguishes two layers of digital infrastructure: hardware-related inflows, proxied by HS 847150 imports, and hosting-related footprint, proxied by secure internet servers. Fixed-effects models show that property registration cost and grid loss are negatively associated with hardware-related inflows, while hosting footprint is more closely tied to broader development level. Middle-income countries face weaker baseline land, grid, and governance conditions, but weaker marginal responsiveness to property-cost changes. The paper proposes infrastructural poverty as a conceptual frame for understanding digital infrastructure absorption constraints in structural transformation.

This study examines the impact of agricultural productivity and governance on structural transformation in selected Sub-Saharan African economies from 2000 to 2023. Using a balanced macro-panel of 12 countries and fixed-effects models with Driscoll–Kraay standard errors, the analysis assesses how productivity gains and institutional quality affect the shares of agricultural employment. Descriptive patterns indicate a steady decline in agricultural employment, accompanied by modest increases in cereal yields, food production, and governance indicators. Empirical results show that higher agricultural productivity, particularly in food production, significantly reduces agricultural employment, consistent with the theory of structural transformation. Governance exerts a negative effect on agricultural labor shares once cross-sectional dependence is accounted for, suggesting that stronger institutions facilitate labor movement out of agriculture by improving incentives and market functioning. The findings highlight the importance of productivity improvements and governance reforms in accelerating structural transformation and supporting policies that strengthen land governance and economic opportunities.

Creating socially and economically developed societies is the ultimate aim of socioeconomic policy, and for nearly 70 years European countries have promoted social innovation (SI) as a primary means of achieving it. Yet whether this investment translates into economic growth remains an open empirical question, largely due to the absence of a proper macroeconomic conceptualization. We address this gap through three contributions. First, we develop a macro SI conceptual framework grounded in Sen and Nussbaum's capabilities approach. Second, we construct a Social Innovation Capabilities Index for 78 countries over 2009–2018 using a Bayesian spatial factor model. Third, using System GMM and sequential G-estimation, we show that SI capabilities drive sustained economic growth — measured by economic complexity — rather than short-run income. We further identify inclusive human capital accumulation as the key mechanism. SI's path to sustained growth is a matter of inclusion and technique, and persists across diverse spatiotemporal contexts.

This paper examines the impact of Zimbabwe’s Fast Track Land Reform Programme (FTLRP) using granular geospatial and household data spanning the pre- and post-reform period. Using a difference-in-difference framework, we compare outcomes on resettled A1/A2 land with communal areas that remained outside the scope of the reform. We also compare Zimbabwe’s agricultural outcomes with those of Zambia, a neighbouring country that shares a similar colonial history, climate, and agro-ecological conditions but did not implement large-scale land reforms during this period. Our results show substantial declines in agricultural performance for most major crops, a slight decline in the household wealth index, and an increase in the share of stunted children. Together, these results suggest that the land reform worsened agricultural productivity and did not deliver the anticipated benefits of improved welfare.

Uganda’s refugee policy is among the most progressive globally, granting new arrivals access to services, freedom of movement, and, when available, small plots of cultivable land intended to foster self-reliance. Yet little empirical evidence exists on whether these land allocations actually improve refugee livelihoods. Using four waves (2017–2021) of the representative RIMA panel, covering refugee and host households across 12 settlements, we estimate the causal impact of initial land allocation on welfare and production outcomes. To address endogeneity in settlement parcel assignments, we implement a leave-one-out instrument based on cohort-level average allocations. Our results show that households with larger initial allocations substantially improve food security, increase agricultural earnings, and raise the share of production sold. Effects are consistently larger in IV than OLS estimates, suggesting that households granted more land may be less prepared to utilize it—consistent with qualitative interviews documenting feelings of limited agricultural experience among new arrivals.

We examine how land ownership shapes educational mobility in rural India. Using full-count rural census microdata, we document a robust step-function pattern across the land distribution: educational mobility rises sharply from the landless to marginal landholders and then plateaus. Exploiting historical variation in British-era land-tenure regimes, we demonstrate a causal link between higher landlessness and lower educational mobility. To unpack mechanisms, we develop a model where parents allocate children’s time between school and work under a subsistence constraint. With little or no land, the constraint binds, increasing child labour and suppressing schooling; a small rise in land relaxes it, producing a sharp drop in child labour and a jump in schooling and upward mobility. The framework endogenously generates the step-function, matches the core facts, rationalizes heterogeneities, and yields testable predictions that we validate.

This study examines the drivers and consequences of household transitions from farm to non-farm livelihoods in Rwanda. Using nationally representative EICV data, we apply Random Forest with SHAP values to identify key determinants of livelihood choice and a Multinomial Endogenous Treatment Effects model to assess impacts on poverty, productivity, and diversification. Results show that households specializing solely in non-farm activities see modest poverty reductions but no significant productivity gains. In contrast, those combining farm and non-farm work achieve higher labor productivity but only marginal poverty reduction. These findings suggest that non-farm work alone does not ensure escape from multidimensional poverty without improvements in job quality and earnings. Meanwhile, diversification enhances resilience but not income per hour. Policy should focus on improving non-farm job quality, supporting mixed livelihood strategies through credit and skills development, and fostering synergies between sectors to ensure productivity gains lead to inclusive rural development.

Literature on the governance of common-pool resources predicts that consensus on resource use rules will promote effective governance of natural resources. However, empirical evidence directly testing this hypothesis is scarce. We investigate how community-level consensus on spatial land use rules is associated with forest conservation using original survey and participatory mapping data from 70 Indigenous communities in the San Martín region of Peru. We conducted an innovative mapping exercise in which community participants independently mapped their understanding of permissible land uses across their territory. We found that (i) areas with higher consensus on forest conservation exhibit significantly lower deforestation rates, while (2) areas with stronger consensus on agricultural use face greater forest loss. Our findings highlight how consensus about the rules governing land use is strongly associated with forest outcomes. This finding suggests that interventions should be deployed to help communities achieve consensus about how to use communal lands.

We provide the first causal evidence on the environmental consequences of secure land property rights by examining the impact of China’s national land titling reform (2009–2019) on biodiversity conservation. Exploiting the staggered rollout of the reform and rich county-level panel data, we implement a difference-in-differences design to identify its effects. We find that land titling significantly improves ecological outcomes, including habitat diversity, bird species richness, crop variety, and ecosystem productivity. Mechanism analyses show that land titling promotes biodiversity by realigning private incentives—through more efficient land use and sustainable investment—and by enhancing community coordination for collaborative conservation activities. By demonstrating how private property rights can help deliver global public goods, our study uncovers a novel institutional channel linking tenure security to environmental sustainability.

Collective land titling in the tropics often reduces deforestation, yet the mechanisms behind this effect remain unclear. Leveraging a unique parcel-level dataset from Bolivia's recent land reform, I show that collective land titles protected forests, especially primary forests, in lowland Bolivia. I investigate four potential mechanisms: low land quality, pre-titling land clearing, restrictive land rights, and collective governance. I find no support for the first two: collective land was in fact more suitable for agriculture than large private land, and collective titles showed no clear pre-trends of deforestation before titling. The evidence instead supports the latter two mechanisms: restrictive land rights appear to limit agricultural expansion in collective titles, and collective titles granted to long-established communities exhibit stronger forest protection compared to those granted to newly formed communities.

This paper examines the impact of the Forest Rights Act (FRA) on the dietary diversity of India's indigenous communities (Schedule Tribes/STs). Historically dependent on forests for subsistence, livelihoods, and cultural identity, STs lacked formal rights until the FRA was enacted in 2008, granting access to forest land and non-timber forest products (NTFPs). Using four rounds of a large-scale consumer expenditure survey and variation in forest cover to capture the potential of the Act, we employ a difference-in-differences strategy. Post-FRA, dietary diversity of ST households increased, driven by an increase in the diversity of vegetables, fruits, and oils consumed. FRA titles data further supports this claim. We document a shift in food sources from subsistence-based collection and cultivation to market purchases. Gains are strongest in areas with larger shares of moderately dense forests that enable NTFP access. Suggestive evidence indicates a shift towards non-agricultural employments potentially facilitated by improved resource access.

Forests generate important public goods, including carbon storage and habitat for biodiversity, while also contributing to increasing wildfire risks. Private landowners do not fully internalize these benefits and risks, potentially leading to socially inefficient management and undersupply of public goods. Exploiting the historical ‘checkerboard’ land pattern in the American West as an instrument for ownership, we show that public ownership reduces timber harvests and herbicide use in forests, resulting in greater carbon storage, higher biodiversity, and, surprisingly, also lower wildfire risk. These results underscore the role of land ownership in shaping the supply of public goods.

This paper develops a novel empirical framework to quantify the social costs of infrastructure development in India, focusing on displacement from large-scale energy projects—coal, hydro, and solar—over the past 50 years. Combining over 25,000 land acquisition notices with satellite-derived settlement data, I construct the first spatio-temporal estimates of project-affected persons (PAPs). Results show displacement varies sharply across sectors, with coal mining imposing the highest burden, particularly on tribal populations in central and eastern India. Displacement from hydro projects peaked in the 1970s-1990s, while coal-related displacement has surged in recent decades. Solar projects, though less intensive, increasingly threaten common lands. Displacement intensity explains nearly 40% of project cost overruns. This methodology enhances the integration of social costs into infrastructure planning and contributes to debates on equitable energy transitions. Without addressing compensation gaps, legal fragmentation, and grievance failures, energy infrastructure push risks perpetuating displacement and deterring future investments.

The sustainability of global agriculture is fundamentally challenged by a nexus of interconnected crises comprising labor scarcity, environmental degradation, and food security. However, the existing market mechanisms and public policies aimed at addressing these challenges have long been constrained by a fundamental “measurement gap”. On the one hand, as a cornerstone of traditional agricultural economics, farm survey data is of fundamental importance. Yet, this data collection method suffers from several systematic flaws, including recall bias, errors in measuring key variables (such as land area), and infrequent sampling (usually annual). For example, research have found that smallholders often report plot sizes that are up to five times larger than the actual size. Such low-quality data not only distorts our understanding of farm productivity and behavior but also undermines the efficacy of policy interventions.

How do social norms shape the intergenerational transmission of land, particularly regarding gender equity? We combine experimental and administrative data from Senegal with a model of social interactions to study how parental preferences and community norms influence land allocation to sons and daughters. A choice experiment that randomizes land allocations and normative cues reveals that parents are more responsive to community norms when allocating land to daughters—a pattern subsequently borne out in administrative records. Simulations show that this asymmetry generates multiple equilibria in land allocation to daughters but not sons, so even modest shifts in perceived norms can meaningfully reduce gender gaps.

Land restoration activities (LRA) on private smallholder farms may reduce land and labor productivity contemporaneously, even though both may increase over time. This temporal tradeoff could reduce smallholder well-being contemporaneously, requiring short-term incentives to compensate for these negative effects. We take the case of Ethiopia, which has the largest land restoration program in the world, to examine the impacts of LRA. We use three panels of the LSMS-ISA data between 2011 and 2016, and remotely sensed data between 2011 and 2018. We find that, contemporaneously, LRA reduces land productivity and labor productivity, but improves household consumption expenditure due to LRA-implementing households working in and receiving transfers from social welfare programs. LRA increases land productivity around five years after implementation. As the adoption of LRA can contemporaneously reduce land and labor productivity, providing short-term social protection through food/cash-for-work programs can encourage the adoption of LRA by reducing rural distress.

Carbon markets are expanding into indigenous lands, but the allocation of carbon rights remains unresolved. In the Philippines, the Indigenous Peoples’ Rights Act recognizes ancestral domains as private community property, yet it is unclear whether carbon value follows that ownership or is shaped by project design. This paper examines how carbon governance is structured in practice and whether it aligns ownership, control, and incentives. Combining doctrinal legal analysis with institutional economics, it analyzes three projects within ancestral domains. The findings show a recurring pattern: despite formal land ownership, control over carbon assets is often exercised through developer-led contractual and registry systems. This separation affects benefit distribution, accountability, and long-term stewardship. Drawing on property rights theory, the paper argues that carbon benefits should presumptively follow ownership and proposes an ownership-based design in which Indigenous communities are default rights holders, and developers operate under delegated authority, strengthening incentives and improving governance outcomes.

Attracting private-sector investment in agriculture is crucial for scaling soil health and climate action, and sustainably transforming food systems. This study assesses how different funding models shape the economic viability and inclusiveness of carbon farming projects. Using evidence from twelve Kenyan projects registered under the voluntary carbon market, four funding models were identified: (1) donor-financed, (2) investor-financed with forward purchase agreements, (3) private-sector led with blended finance, and (4) buyer-led. The donor-financed model enables piloting but risks donor dependence and high costs. Investor financing offers stable funding yet limits future gains. Private-sector models allow marketing flexibility but face financial risks, while buyer-led models centralize power and may weaken equitable benefit-sharing. Four cross-cutting insights arise: investment terms shape risk–reward distribution; early scaling and profit reinvestment support economic viability; fair benefit-sharing builds on transparency in cost accounting; and inclusive governance requires empowerment of local actors and reduced reliance on international intermediaries.

This paper examines the impact of land rights on Land use or investments in Uganda. I used a sample of 3548 households from the 2019/2020 Uganda National Household Survey (UNHS). Analytically I employed maximum likelihood probit techniques. land rights is expressed in three forms: land tenure, selling rights, and ownership of land title certificate.

The study found that; mailo tenure, selling rights, and title ownership encourage coffee cultivation. Freehold and mailo tenure system support the growing of trees for timber. All the three types of land rights encourage investing in permanent houses. Following the above findings, we recommend the need for government to facilitate households acquire title ownership to enhance land investments.

We explore the effect of historical ethnic borders on contemporary conflict in Africa. We document that the intensive and extensive margins of contemporary conflict are higher close to historical ethnic borders. Exploiting variations across artificial regions within an ethnicity's historical homeland and a theory-based instrumental variable approach, we find that regions crossed by historical ethnic borders have 27 percentage points higher probability of conflict and 7.9 percentage points higher probability of being the initial location of a conflict. We uncover several key underlying mechanisms: competition for agricultural land, population pressure, cultural similarity, and weak property rights.

We examine the relationship between agricultural land potential and conflict across Sub-Saharan Africa, where such conflicts are widespread. To address endogeneity, we construct a plausibly exogenous, crop-specific indicator of agricultural land potential based on a weather-driven suitability model. The indicator leverages variation in weather and soil conditions to capture changes in crop-growing potential. Using a 0.1° grid-cell dataset covering the period 1997–2023 and conflict data from ACLED, we focus on overall crop and climate suitability for 50 crops grown across Africa. We find that long-term increases in agricultural land potential are associated with increases in conflict within countries. By using crop-specific, weather-based suitability measures rather than realized production, we provide new evidence on the climate–agriculture–conflict nexus.

We examine how armed groups and state authorities shape forest governance in a post-conflict setting. Using an original household survey (N = 2,024) and satellite data from the Colombian Amazon, we find that armed groups impose more rules, monitoring, and sanctions than other actors. However, forest loss is higher in areas under their control, indicating that rule proliferation does not improve conservation outcomes. In contrast, subnational governments actively combat deforestation, and their presence is associated with lower forest loss. These effects are strongest among households engaged in illegal activities, particularly coca cultivation and cattle ranching. The findings reveal a central paradox of post-conflict environmental governance: armed groups generate extensive regulatory systems that fail to curb deforestation. Overall, the results clarify the environmental role of armed groups in contexts with strong incentives for illegal production and ongoing competition between state and non-state authorities.

This paper examines how formalizing women’s land rights improves adoption of climate‑smart agricultural practices and strengthens household resilience in eastern Democratic Republic of Congo. Drawing on mixed‑methods research from a three‑year program implemented by Nature Seed Care Agri‑Organization, I combine household surveys (n=1,200), georeferenced plot data, and 40 qualitative interviews with women farmers, local leaders, and extension agents. Results show that women with legally recognized land use or tenure arrangements are 35% more likely to adopt soil conservation, agroforestry, and diversified cropping systems, and exhibit higher seasonal food security scores. Qualitative evidence highlights barriers to title registration, including administrative costs, customary norms, and limited access to legal services, and identifies community‑led mediation and mobile registration outreach as effective solutions. The paper discusses policy implications for scaling gender‑responsive land formalization, integrating tenure security into climate finance, and designing extension services that accelerate equitable, climate‑resilient agricultural uptake.

Using linked property transaction data and credit records, we document group-level differences in financial distress before and after transitioning into homeownership. Black and female borrowers experience larger increases in post-purchase delinquency relative to White and male borrowers, respectively, as measured by 30+ day and 90+ day delinquency rates, even after accounting for stringent controls, credit score, and income-at-purchase. Post-purchase, Black and female borrowers also accumulate more student debt, while credit card balances rise for everyone, but faster for Black borrowers. Understanding disparities in sustaining homeownership informs our assessment of racial and gender disparities in the financial benefits of homeownership.

We identify and quantify a novel gender gap in the housing market for singles in South Asia. Using nationally representative data from India, we document that single women face a housing cost premium of 3-4% relative to men. We leverage detailed data from India's two largest online property listing portals, covering more than 230 cities, we find that shared housing accommodation is listed for rent at a rate that is 6-8% higher for women than for men, after controlling for an extensive set of housing arrangements, amenities, and granular neighborhood fixed effects. We find similar gendered housing premiums in property listings data from Bangladesh and Sri Lanka. Gaps persist for (gendered) accommodation that is offered within the same building, but vanishes for owners that have listings for both men and women. Variation between cities in this gender gap can be explained by historical gender norms.

Land pooling and readjustment (LR) schemes aim to secure land for public amenities including affordable housing, schools, and healthcare centres while enabling coordinated private development. In Ahmedabad, however, the delivery of public-use plots varies widely: some are fully developed, while others remain vacant or occupied by informal settlements. This study examines why land allocated for public purposes is often not transferred to or developed by the municipal government. Qualitative fieldwork reveals an incentive structure in which officials prioritize smooth private development, leading to the placement of some public-use plots on parcels with disputed or incomplete property rights. Such plots are subsequently difficult for the state to acquire and develop. A citywide quantitative model tests these findings using a stratified sample of public-use plots combined with spatial, administrative, and ownership-conflict data. The study provides new evidence on how institutionalized incentives shape the effectiveness of LR in delivering essential public amenities.

I exploit Delhi's 2014 Floor Area Ratio reform, which raised allowable FAR by 33-60% for residential parcels exceeding 750 square meters, to examine whether relaxing land-use regulations improves housing affordability in developing countries. Using satellite-derived building heights at 4m resolution, administrative parcel boundaries, and proprietary housing data in a sharp regression discontinuity design, I find economically negligible height increases of 0.58-0.74 meters (3-14%) nine years post-reform, with no spillovers to neighboring parcels. Despite 306% nominal price increases, the formal housing market exhibited no significant supply expansion. Peripheral areas showed treatment effects four times larger than central zones, inverting standard bid-rent gradients. I develop a structural model demonstrating that upzoning can fail when developers face convex construction costs, favor horizontal fragmentation, and are forward-looking. These findings suggest relaxing land-use regulations is necessary but insufficient for addressing housing affordability in contexts with coordination failures, institutional frictions, and weak property rights.

In many low to middle income countries, around 70% of land remains undocumented, unmanaged, or unprotected by the state, including in peri-urban and urban areas. The Sustainable Development Goals have established targets to monitor tenure security, prompting the development of initiatives that assess land rights and governance through data collected from experts, practitioners, community representatives, and residents. However, how these toolkits measure land rights and governance—and the extent to which their approaches are comprehensive—remains unclear. This paper provides a systematic critical review of toolkits developed by international organizations and other land rights stakeholders, compiled in the MetRe Library online (at this link). It offers an unprecedented synthesis of empirical approaches to studying de facto land rights and social regulations, while also identifying gaps and proposing improvements. In doing so, the research supports the development of more robust and comprehensive tools to inform both academic inquiry and effective land policy.

Land governance reforms are crucial for inclusive investment and economic resilience, yet data fragmentation and institutional inefficiencies remain major barriers. This study investigates how integrating Financial Technology (FinTech) with administrative land records, survey data, and remote sensing can strengthen land governance in Freetown, Sierra Leone. Using a Sequential Mixed Method design, a quantitative survey of 400+ respondents (analytical sample = 384; Krejcie & Morgan, 1970) is followed by qualitative interviews for explanatory depth. Structured questionnaires measure digital finance usage (independent variables), institutional trust (mediator), and inclusive investment behavior (dependent variable). Administrative cadastral data and satellite imagery spatially anchor transaction and land use patterns. Semi-structured interviews with land administrators, financial providers, and communities explore institutional bottlenecks and FinTech adoption. Econometric modeling, mediation analysis, and spatial diagnostics assess FinTech–investment linkages. The study provides methodological and empirical insights, offering evidence-based strategies for governments and development partners to drive equitable economic transformation.

This first global report on the Status of Land Tenure and Governance brings together tenure- and sex-disaggregated data and analyses from a wide range of sources - including governments, civil society and academia - and across multiple levels, from local to global, with the objective of tracking progress with regards land tenure systems and governance frameworks. It serves policymakers, intergovernmental organizations, civil society, the private sector and academia as a clear and authoritative reference point on land tenure and governance data and analysis. The report supports progress toward numerous SDGs, promotes the uptake of the VGGTs, and contributes to the implementation of other key international frameworks, including the CFS Principles for Responsible Investment in Agriculture and Food Systems (CFS-RAI) and the Rio Conventions.

Forest-governance effectiveness is increasingly linked to gender inclusion, yet cross-national, project-level quantitative evidence remains limited. This study leverages Large-Language Models (LLMs) to overcome these data constraints and provides new empirical evidence on the role of gender policies in improving forest-governance outcomes. We utilize an LLM-based key information extraction (KIE) framework to extract cross-country, project-level variables by processing World Bank REDD+ project documents. We then assess the effectiveness of gender-inclusion strategies—including design-stage and monitoring and evaluation (M&E) stage—on project performance using a Bayesian linear mixed model. Our findings highlight that gender inclusion at the design stage does not improve subjective or objective outcomes, whereas M&E stages enhance not only governance processes but also project outcomes. This work demonstrates the capacity of LLMs to support quantitative policy and project evaluation and, for the first time, extends their application to outcome-oriented research on gender inclusion policy.

Land grabbing in the Colombian Amazon is intricately linked to historical patterns of territorial control and ongoing landscape transformation. Land use change connected with deforestation, conversion to pastures and ultimately land grabbing behaves differently according to historical, social and ecological context. For that purpose, using deforested patches between 2020 and 2021 and the 26 variables related to environmental social, economic, and armed conflict conditions in the Northwestern Arc of the Colombian Amazon, 5 areas were classified using the unsupervised classification algorithm "K-means". In each zone, the similarity between the average values of the variables associated with the patches and their geographic proximity allows for quantitative outputs and geographic representation and facilitates analyzing forest cover loss and local conditions to demonstrate land grabbing in relation to the conversion to pastures and deforestation.

This article analyzes the integration between Brazil’s Rural Environmental Registry (CAR) and the Land Management System (SIGEF) in the Legal Amazon. Employing geospatial similarity (Jaccard Index) and ownership checks using taxpayer identifiers (CPF and CNPJ), it assesses how interoperability can reduce boundary inconsistencies and duplicate records, thus improving environmental enforcement and land governance. The September 2025 national CAR–SIGEF assessment covers nine Legal Amazon states, revealing that 31.57% of CAR registrations are spatially linked to SIGEF parcels, and 91.65% of SIGEF parcels are referenced by a CAR record. The article discusses recent federal initiatives such as the Pre-Filled CAR, PRONAF for Land Regularization, and the Meu Imóvel Rural platform. Recommendations include institutionalizing interoperability standards, increasing support for small producers and traditional communities, and strengthening ongoing validation and monitoring to enhance the integrity and efficiency of land management in the region.

How can weak bureaucracies implement policies? I investigate Brazil’s underresourced land agency as it attempts to formalize property rights for smallholders and traditional communities in Amazonia—a necessary step for sustainable development and halting deforestation. A comparison of neighboring jurisdictions with divergent outcomes reveals counterintuitive conditions under which underresourced organizations can nonetheless function effectively. Ethnography and process tracing show that empowered but fragmented civil society organizations and clashing elites polarized bureaucrats and eroded organizational capacity in one jurisdiction. Conversely, a weak civil society and uninterested elites provided de facto autonomy, enabling moderately partisan bureaucrats to collaborate with clients. Protests and litigation were insufficient to implement policies, while cooperative labor by clients facilitated legibility. Employing a relational political economy approach to analyze state performance, the paper argues that underresourced bureaucracies require shelter from, rather than active support by, elites to operate effectively.

Does industrial activity lead to deforestation and land degradation? Can limited state capacity be overcome to decouple economic activity from environmental degradation? We study these questions in the context of slaughterhouse plant openings in Brazil. Using difference-in-differences, we show that opening a plant increases livestock production while reducing forest cover and degrading pastureland. However, after the introduction of legally-enforceable commitments between slaughterhouses and federal prosecutors that penalize plants for buying livestock from illegally deforested areas, opening a plant leads to higher productivity without increasing deforestation. Our results suggest that incentive-compatible commitments can generate positive economic and environmental outcomes.