When Does Agroforestry Income Reduce Deforestation? Evidence from a Natural Experiment in Madagascar
Camille DeSisto Sustainability Institute, Rice University, Houston, TX, USARanaivo RasolofosonSchool of the Environment, University of Toronto, Toronto, CanadaMichelle FoleyMacMillan Center for International and Area Studies,Yale University, New Haven, CT, USA, Harsh Parikh Yale School of Public Health, Yale University, New Haven, CT, USA, [6pt] Corresponding author: [email protected] · 2026
Read When Does Agroforestry Income Reduce Deforestation? Evidence from a Natural Experiment in Madagascar by Camille DeSisto Sustainability Institute, Rice University, Houston, TX, USARanaivo RasolofosonSchool of the Environment, University of Toronto, Toronto, CanadaMichelle FoleyMacMillan Center for International and Area Studies,Yale University, New Haven, CT, USA, Harsh Parikh Yale School of Public Health, Yale University, New Haven, CT, USA, [6pt] Corresponding author: [email protected] (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-13706.txt.
Introduction
Camille DeSisto and co-authors ask when agroforestry income reduces deforestation, using a natural experiment in Madagascar. The 2015 global vanilla price boom gave an exogenous income shock to Madagascar's main vanilla region. With a matching-augmented synthetic control design, they estimate income gains cut annual deforestation in 2017 by a set margin. Effects diverged by place, though: income reduced clearing in drier, accessible municipalities but raised it in wetter, low-elevation areas. Income both relaxes subsistence pressure and raises the opportunity cost of conservation, they argue.
Catalog record
| CREATOR | Camille DeSisto Sustainability Institute, Rice University, Houston, TX, USARanaivo RasolofosonSchool of the Environment, University of Toronto, Toronto, CanadaMichelle FoleyMacMillan Center for International and Area Studies,Yale University, New Haven, CT, USA, Harsh Parikh Yale School of Public Health, Yale University, New Haven, CT, USA, [6pt] Corresponding author: [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46602 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.13706
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
When Does Agroforestry Income Reduce Deforestation? Evidence from a Natural Experiment in Madagascar Camille DeSisto Sustainability Institute, Rice University, Houston, TX, USARanaivo RasolofosonSchool of the Environment, University of Toronto, Toronto, CanadaMichelle FoleyMacMillan Center for International and Area Studies,Yale University, New Haven, CT, USA, Harsh Parikh Yale School of Public Health, Yale University, New Haven, CT, USA, [6pt] Corresponding author: [email protected] Abstract Tropical deforestation and rural poverty are deeply intertwined, yet isolating the causal effect of income on forest loss remains challenging. We use the 2015 global vanilla price boom—triggered by food-industry shifts toward natural flavoring—as an exogenous income shock affecting Madagascar’s primary vanilla-producing region. Using a matching-augmented synthetic control design, we estimate that income gains reduced annual deforestation by 1.7 percentage points in 2017, equivalent to approximately 701 hectares of avoided forest loss. Under a monotonicity assumption linking the price boom to farmers’ income, the sign of this reduced-form effect is informative about the causal direction of income on deforestation. However, effects were strongly heterogeneous: higher incomes reduced deforestation in drier, more accessible municipalities but increased clearing in wetter, low-elevation areas with high agricultural potential. These divergent patterns suggest that income simultaneously relaxes subsistence pressures driving forest dependence and raises the opportunity cost of conservation where agricultural returns are high. Our findings indicate that commodity-based agroforestry can align poverty alleviation with forest conservation under conditions of low agricultural opportunity cost. Still, policies must anticipate contexts where rising incomes amplify deforestation in agriculturally suitable land. The strategic targeting of livelihood interventions based on local agricultural potential may help reconcile development and conservation objectives in tropical forest frontiers. Preprint — Not peer reviewed August 10, 2026 1 Introduction Commodity crops sit at the center of tropical land-use change ( 23 ) . Since the 1980s, export-oriented agriculture has expanded rapidly as smallholder, state-enabled clearing gave way to enterprise-driven deforestation ( 33 ; 55 ; 14 ) . Converting forests to commodity production—soybeans, palm oil, cocoa, and others—drives carbon emissions and soil carbon loss, perturbs regional rainfall and land-surface temperature, and elevates infectious disease risk ( 48 ; 50 ; 60 ; 38 ; 12 ) . Economic theory and a large empirical literature indicate that higher commodity prices increase deforestation by raising returns to land conversion in open economies ( 3 ; 11 ; 7 ; 64 ; 18 ; 20 ; 4 ; 37 ) . Yet, for rural households in low- and middle-income countries, the same crops provide crucial cash income and buffer shocks. Agroforestry—the integration of agriculture and trees/shrubs—may reconcile poverty alleviation and forest conservation goals. Commodity-based agroforestry offers relatively high earnings while retaining tree cover and ecosystem functions ( 56 ) ; income support and alternative livelihoods can reduce reliance on risky, often illegal forest clearing ( 45 ; 15 ; 31 ; 59 ) . The net effect of a commodity price shock on deforestation is therefore ambiguous and context-dependent: the same price increase that raises the return to land conversion also raises household income, potentially reducing pressure on forests through substitution away from more destructive livelihoods. Figure 1 : Vanilla cultivation in the agroecological matrix of rural northeast Madagascar. A farmer hand-pollinates vanilla orchids grown on tutor trees, reflecting the labor-intensive nature of the crop. Deforestation in the rural tropics is often a dangerous, last-resort activity driven by economic necessity. If higher incomes relax this pressure, wealth should reduce deforestation. However, the same income gains can incentivize forest clearing to expand agricultural production, making the net effect ambiguous. We use the 2015 global vanilla price boom—an exogenous shock to regional income in Madagascar’s SAVA region—to isolate this relationship. On average, the income shock reduced deforestation. However, effects were strongly heterogeneous: income decreased deforestation in drier, more accessible municipalities and increased it in wetter, low-elevation areas with high agricultural potential. Madagascar & Vanilla. Madagascar vanilla exemplifies this tension. Madagascar is a biodiversity hotspot with persistent forest loss and pervasive rural poverty ( 43 ; 62 ; 58 ; 65 ) . Madagascar vanilla ( Vanilla planifolia ) is a high-value export that supports hundreds of thousands of livelihoods and a substantial share of national exports ( 2 ; 24 ; 66 ) . The majority of the world’s vanilla is grown in the SAVA region of northeast Madagascar, where approximately 80% of households farm vanilla ( 35 ; 34 ) . Although the SAVA is wealthier than many other regions in Madagascar, its residents remain burdened with high levels of poverty and food insecurity ( 34 ; 27 ; 26 ) . The natural history of vanilla renders its cultivation labor-intensive and often risky, but potentially complementary to forest conservation. Vines reach maturity approximately three years after planting. Because V. planifolia is native to Mexico, each flower must be hand-pollinated to overcome the lack of co-occurring pollinators in Madagascar. Post-harvest curing and drying further expand the workload, and farmers often experience crop loss due to violence, pathogens, and cyclones ( 63 ; 49 ; 30 ; 25 ; 22 ) . Unlike other tropical commodity crops (e.g., palm oil), vanilla is an orchid best cultivated using tutor trees for shade and structural support; grown under agroforestry settings, it can complement biodiversity conservation goals ( 24 ; 66 ) . In some cases, however, rainforest is cleared for vanilla cultivation, directly contributing to deforestation. A vanilla price shock could therefore exacerbate or mitigate deforestation through multiple channels. In 2015, Madagascar vanilla experienced a price boom triggered by shifts in the food-industry towards natural flavoring ( 46 ; 51 ) . We treat the 2015 vanilla price boom as a natural experiment to investigate the effects of income on deforestation. Anecdotally, following the 2015 boom, farmers rebuilt homes, bought new land, and started businesses selling furniture ( 57 ) —activities that likely increased demand for forest-sourced planks and spurred land clearing. However, higher vanilla returns could also turn farmers away from more destructive practices: high returns may disincentivize lower-value crops (e.g., rice, often grown via swidden clearing, or tavy ) and logging in difficult-to-access forests. Although vanilla-producing areas have shown lower historical deforestation ( 42 ) , such associations are not necessarily causal. Using the 2015 price boom as a natural experiment, we therefore ask: how does the vanilla price boom affect deforestation in vanilla-producing municipalities, and what does this reveal about the relationship between agricultural income and forest loss? Study Design & Findings. We use the 2015 vanilla price boom as a plausibly exogenous shock, comparing deforestation trajectories in Madagascar’s 73 vanilla-producing SAVA municipalities (treated) with matched non-vanilla municipalities (controls) using a matching-augmented synthetic control design. Under a monotonicity assumption linking the price boom to farmer income, the sign of our reduced-form estimates is informative about the causal direction of income on deforestation (Lemma 1 ). On average, the price shock reduced deforestation by 1.7 percentage points in 2017 ( {\sim}701 ha of avoided forest loss), but effects were strongly heterogeneous: deforestation declined in drier, more accessible municipalities and increased in wetter, low-elevation areas with high agricultural potential. These patterns are consistent with income simultaneously relaxing subsistence pressures and raising the opportunity cost of conservation where agricultural returns are high. 2 Results Study context. From 2005–2012, deforestation rates