DETERring more than Deforestation: Environmental Enforcement Reduces Violence in the Amazon (2509.06076v1)

Abstract: We estimate the impact of environmental law enforcement on violence in the Brazilian Amazon. The introduction of the Real-Time Deforestation Detection System (DETER), which enabled the government to monitor deforestation in real time and issue fines for illegal clearing, significantly reduced homicides in the region. To identify causal effects, we exploit exogenous variation in satellite monitoring generated by cloud cover as an instrument for enforcement intensity. Our estimates imply that the expansion of state presence through DETER prevented approximately 1,477 homicides per year, a 15% reduction in homicides. These results show that curbing deforestation produces important social co-benefits, strengthening state presence and reducing violence in regions marked by institutional fragility and resource conflict.

• The paper finds that each additional deforestation-related fine reduces homicide rates by 0.73 per 100,000 inhabitants.
• It employs a 2SLS approach using cloud cover as an instrument to establish a causal link between enforcement intensity and violence reduction.
• The study estimates that robust enforcement prevented nearly 1,477 homicides annually, highlighting substantial social co-benefits.

Environmental Enforcement and Violence Reduction in the Brazilian Amazon

Introduction

This paper provides a rigorous empirical analysis of the causal impact of environmental law enforcement on violence in the Brazilian Amazon. The authors exploit the implementation of the Real-Time Deforestation Detection System (DETER), a satellite-based monitoring system that enables rapid detection and sanctioning of illegal deforestation. The central hypothesis is that intensified environmental enforcement, by increasing the cost of illegal land use and resource extraction, can reduce violent conflict in regions characterized by institutional fragility and contested land tenure. The paper leverages exogenous variation in satellite monitoring capacity induced by cloud cover as an instrument for enforcement intensity, enabling credible identification of causal effects.

Institutional Context and Theoretical Framework

The Amazon region is marked by weak state presence, insecure property rights, and pervasive illegal markets. These conditions foster strategic violence as a mechanism for contract enforcement, dispute resolution, and territorial control. Unlike the urban-centric pattern of violence in most of Brazil, the Amazon exhibits a rural configuration, with lethal violence concentrated near deforestation frontiers and zones of illegal mining and logging.

Deforestation in the Amazon is predominantly illegal, driven by land grabbing, agricultural expansion, and resource extraction. The institutional response, notably the PPCDAm and the DETER system, represents a shift toward integrated, proactive environmental governance. DETER's real-time satellite alerts facilitate targeted enforcement, increasing the perceived probability of detection and sanction.

The authors formalize the ambiguous effect of enforcement on violence using an occupational choice model. Individuals select among legal work, deforestation, and crime, with payoffs determined by returns and detection probabilities. The model demonstrates that enforcement can either reduce or increase violence, depending on the relative propensity of deforestation and crime to generate conflict and the returns to each activity.

Data and Descriptive Analysis

The empirical analysis utilizes a municipality-year panel (2006–2016) covering 521 Amazon biome municipalities. Key variables include:

• Homicide rate (SIM-DataSUS, IBGE): deaths per 100,000 inhabitants, with inclusive ICD-10 coding to mitigate underreporting.
• Environmental enforcement (IBAMA): number of deforestation-related fines issued.
• Cloud coverage (INPE-DETER): proportion of municipal area obscured, serving as an instrument for enforcement intensity.
• Controls: satellite observability, weather (precipitation, temperature), and socioeconomic indicators (commodity prices, GDP, population density, education quality).

Spatial heterogeneity is pronounced, with substantial variation in violence, enforcement, and cloud coverage across municipalities.

Figure 1: Geographical distribution of homicide rates, environmental enforcement, and DETER cloud coverage in 2006, illustrating spatial heterogeneity in key variables.

Empirical Strategy

The main identification challenge is the endogeneity of enforcement: municipalities with higher violence may receive more enforcement resources, and unobserved factors may confound the relationship. The authors implement a two-stage least squares (2SLS) approach, instrumenting enforcement intensity with cloud coverage, which exogenously impairs satellite monitoring and thus enforcement.

The second-stage specification estimates the effect of lagged enforcement (fines issued in the previous year) on homicide rates, controlling for a rich set of covariates and fixed effects. The lag structure accounts for behavioral response delays and mitigates reverse causality.

Main Results

The 2SLS estimates reveal a robust, negative, and statistically significant effect of environmental enforcement on homicide rates. The preferred specification indicates that one additional deforestation-related fine reduces the homicide rate by 0.73 per 100,000 inhabitants (2.58% relative to the mean). An increase from the 25th to the 75th percentile in enforcement intensity (≈8 fines) yields a 20.7% reduction in homicide rates.

Figure 2: Divergence in homicide rates between Amazon and non-Amazon municipalities, highlighting the region's distinctive violence trajectory.

Scaling the effect to the Amazon's population, the authors estimate that DETER prevented approximately 1,477 homicides per year—a 15% reduction from the annual average. The benefit–cost ratio, based solely on violence reduction and willingness-to-pay estimates, is at least 3.7, even excluding environmental objectives.

Robustness and Extensions

The results are robust to alternative homicide definitions, sensitivity analyses for omitted variable bias, and distributional regressions. The sensitivity analysis (Cinelli et al., 2025) demonstrates that an omitted variable would require implausibly large explanatory power to invalidate the causal interpretation. Distributional regressions show that enforcement reduces both the incidence and endemicity of violence: each additional fine lowers the probability of any homicide by 1.2 percentage points and the probability of endemic violence (rate >10) by 1.8 percentage points.

Mechanisms

Three channels plausibly mediate the observed effect:

1. Reduction in deforestation and land conflict: Enforcement curtails illegal clearing, easing disputes over land appropriation.
2. Expansion of state presence and deterrence: Visible enforcement signals institutional control, increasing perceived risks for criminal actors.
3. Disruption of illegal economic activities: Enforcement raises costs and logistical barriers for organized crime linked to land grabbing, logging, and mining.

The empirical strategy does not disentangle these mechanisms, but prior literature supports their simultaneous operation.

Policy Implications and Future Directions

The findings challenge the conventional trade-off between environmental protection and development, demonstrating that environmental enforcement yields substantial social co-benefits in violence reduction. Policymakers should integrate environmental monitoring into broader governance and public safety strategies, particularly in regions with weak institutions.

Future research should address the political economy of enforcement allocation, strategic adaptation by criminal networks, and long-run, intergenerational effects of environmental interventions. The interaction between environmental and other forms of law enforcement, as well as the potential for spillover effects, warrants further investigation.

Conclusion

This paper provides compelling evidence that environmental law enforcement, operationalized through real-time satellite monitoring and sanctioning of illegal deforestation, significantly reduces violence in the Brazilian Amazon. The causal estimates are robust, economically meaningful, and policy-relevant. The multidimensional benefits of environmental enforcement—spanning conservation and public security—should be incorporated into policy evaluation and cost–benefit analysis. The integration of environmental and security objectives is essential for effective governance in frontier regions characterized by institutional fragility and resource conflict.