Main drivers of deforestation

Global and regional evidence

Evidence quality: Moderate–high (FAO, 2022; Curtis et al., 2018; Pendrill et al., 2019). FAO's sample-based remote-sensing assessment classified land-use changes across 2000, 2010, and 2018 (FAO, 2022). It attributed almost 90% of global deforestation during 2000–2018 to agricultural expansion: nearly 50% to cropland expansion and 38.5% to livestock grazing (FAO, 2022). The regional pattern was not uniform; the same assessment attributed almost three-quarters of South American deforestation to livestock grazing (FAO, 2022). These results support the stub's claim that animal agriculture is a major deforestation-linked driver in many, but not all, contexts (FAO, 2022).

A separate global analysis used satellite imagery to classify dominant drivers of tree-cover loss from 2001 to 2015 (Curtis et al., 2018). It estimated that permanent conversion for commodity production accounted for just over one-quarter of tree-cover loss and identified beef, soy, palm oil, and wood fiber among the commodities involved (Curtis et al., 2018). That classification distinguishes permanent conversion from forestry, shifting agriculture, wildfire, and urbanization, which is important because not every instance of tree-cover loss is deforestation (Curtis et al., 2018).

Demand through supply chains

Production location and final demand were separated in a pan-tropical model that traced emissions from forest conversion through agricultural and forestry supply chains for 2010–2014 (Pendrill et al., 2019). It estimated approximately 2.6 gigatonnes of carbon-dioxide emissions per year from the expansion of agriculture and tree plantations into tropical forests; cattle and oilseed products accounted for more than half, and international trade drove an estimated 29–39% of the total (Pendrill et al., 2019). This supports the existing claim that trade-embedded deforestation is materially associated with beef and oilseed supply chains, although the study reported cattle products rather than a combined beef-and-leather category (Pendrill et al., 2019). A Brazilian cattle supply-chain study separately documented beef and leather markets within the same upstream production system and analyzed their exposure to deforestation-linked cattle supplies (Walker et al., 2013).

Exports were not the only demand source in that analysis (Pendrill et al., 2019). Pendrill and colleagues found that cattle meat was more often consumed domestically than oilseed products in producer countries, while a substantial share of oilseed-linked emissions was embodied in trade (Pendrill et al., 2019). Livestock-related demand can therefore appear through grazing expansion, domestic beef markets, exported beef, leather markets, and feed-related oilseed markets, with different combinations across countries (FAO, 2022; Pendrill et al., 2019; Walker et al., 2013).

Mediating conditions

Commodity demand operates alongside access, institutions, and land-use policy (Busch and Ferretti-Gallon, 2017). A meta-analysis of 121 spatially explicit econometric studies found more clearing where agricultural returns were higher or clearing and transport costs were lower; it also found that roads, population, protected areas, law enforcement, and indigenous forest management can affect outcomes, although several governance and tenure associations were inconsistent or based on preliminary evidence (Busch and Ferretti-Gallon, 2017). Governance and land rights are therefore relevant mediators, but the literature does not support one universal effect size for either factor (Busch and Ferretti-Gallon, 2017).