Why Is Deforestation Happening? The Main Causes

Agriculture is, by a wide margin, the single largest reason forests disappear. Between 2001 and 2022, roughly 86% of global deforestation was tied to crop and cattle production.¹ But farming alone does not explain why the chainsaws keep running. Roads punched into wilderness, mining operations, charcoal production, global commodity markets, misaligned government subsidies, and weak land governance all feed the process in distinct ways. The causes vary sharply by region, and they interact with one another in ways that make any single-cause explanation misleading.

Cattle and Soy in Latin America

The Brazilian Amazon is the most studied deforestation frontier on Earth, and the pattern there is consistent: pasture expansion for cattle drives the bulk of clearing. Soybean cultivation plays a secondary but important role, partly by displacing cattle ranchers northward into intact forest. Research has shown that soy expansion in the southern state of Mato Grosso pushed pasture further into the Amazon, causing deforestation far from where the soybeans were actually planted.² Both industries respond to international market signals, including events as seemingly unrelated as mad-cow-disease outbreaks in other countries and surging demand from China.³

Deforestation for cattle persists despite years of public and private efforts to stop it. The properties most likely to clear forest tend to be remote, sell fewer cattle, operate early in the supply chain, and still have a large share of standing forest on their land.⁴ In other words, the problem is concentrated among smaller, harder-to-monitor operations at the edges of the frontier, not necessarily the large industrial ranches that make headlines.

Oil Palm in Southeast Asia

In Indonesia and Malaysia, oil palm has been the dominant driver of forest conversion. Over nineteen years, the area under oil palm in Indonesia doubled to more than 16 million hectares. During that time, forest area shrank by about 11%, and roughly a third of the lost forest was converted directly into oil palm plantations. Industrial plantations were responsible for about three times more forest clearing than smallholder operations.⁵ Much of this expansion took place on carbon-rich and biodiversity-rich land, contributing to record carbon emissions and species loss in the region.⁶

Indonesia’s deforestation rates did eventually slow in the late 2010s, and the timing tracked closely with falling palm oil prices and a slowdown in new plantation development, not just government policy. That connection is a useful reminder that commodity economics often matter more than regulation in determining whether forests stand or fall.

Smallholder Clearing in Africa

The Congo Basin tells a very different story from the Amazon or Southeast Asia. Here, an estimated 84% of forest disturbance comes from small-scale, non-mechanized clearing for agriculture, not from industrial operations. Annual rates of smallholder clearing in primary forests doubled between 2000 and 2014, mirroring population growth. In the Democratic Republic of the Congo alone, smallholder clearing accounted for nearly two-thirds of total forest loss in the basin.⁷

Well-intentioned agricultural development programs can inadvertently worsen the problem. A large-scale experiment in the DRC found that providing smallholder farmers with improved seed varieties led to increased deforestation of primary forest, particularly in remote villages without good road access. Farmers with better cereal seeds appeared to clear more fertile forestland for planting.⁸ The finding complicates a common assumption that agricultural intensification automatically spares forests by raising yields on existing land.

Roads as a Catalyst

A road through intact forest might seem like a minor intrusion, but its downstream effects are enormous. When a “first-cut” road penetrates an untouched forest, secondary roads quickly branch off it like tributaries. In the Brazilian Amazon, researchers found an average of 49 kilometers of secondary road for every kilometer of initial road built. The forest loss and degradation tied to those secondary roads was over 300 times greater than what was directly linked to the first-cut road itself.⁹ The same study found the multiplier effect in the Congo Basin and New Guinea as well, though at lower magnitudes.

Logging roads are a prime example. Loggers build road networks to extract timber, then move on. But the roads remain, and settlers use them to access land for fields and pastures. The networks end up inducing more forest fragmentation than would have been necessary to extract the wood alone.¹⁰ In this way, logging is often less damaging for the trees it removes than for the access it creates.

Mining

Mining’s footprint on forests is smaller than agriculture’s but still substantial and frequently underestimated. A 2022 pantropical assessment found that industrial mining directly cleared about 3,264 square kilometers of tropical forest, with 80% of that loss concentrated in just four countries: Indonesia, Brazil, Ghana, and Suriname. Beyond the mine sites themselves, mining caused indirect forest loss in two-thirds of the countries studied.¹¹

In the Brazilian Amazon specifically, the off-site effects are striking. Mining caused deforestation not just within lease boundaries but across a zone stretching up to 70 kilometers from the mine. Of the roughly 38,000 square kilometers of forest cleared in those zones between 2005 and 2015, about 31% was induced by mining activity.¹² Roads, settlements, and secondary economic activity around mines all contribute to that ripple effect.

A more recent global analysis put the total forested area directly cleared by mining worldwide at nearly 20,000 square kilometers since the start of the century, two to three times higher than previous estimates. About two-thirds of that deforestation was attributed to unrecorded mining activities, meaning operations that do not appear in official mine registries.¹³ Informal and illegal mining, in other words, is a much bigger piece of the picture than official data suggests.

Charcoal and Fuelwood

In sub-Saharan Africa, where hundreds of millions of people depend on wood and charcoal for cooking, energy demand is a distinct and major driver of forest loss. In Mozambique, charcoal production for urban consumers drives forest degradation largely independently of agricultural expansion, and its impact on forest cover is on the same order of magnitude as farming-related deforestation.¹⁴ In Malawi, most firewood and nearly all charcoal comes from indigenous miombo woodlands, the country’s main natural vegetation.¹⁵ This driver rarely makes global headlines but affects vast areas of dry tropical forest and woodland.

Global Trade Amplifies Local Clearing

A significant share of tropical deforestation is not driven by local needs but by international demand for commodities. Depending on the trade model used, somewhere between 29% and 39% of carbon emissions from deforestation are driven by international trade.¹⁶ That share is considerably higher than the proportion of fossil-fuel emissions embodied in trade, which highlights how disproportionately connected deforestation is to global supply chains.

Four commodities stand out: beef, soybeans, palm oil, and wood products. Between 2000 and 2011, production of these four in seven high-deforestation countries was responsible for about 40% of total tropical deforestation. Over a third of the environmental impact was embodied in exports, up from about a fifth at the start of the period. Major flows run from Latin America to Europe, China, and the Middle East (beef and soy) and from Southeast Asia to China, India, and the EU (palm oil and wood).¹⁷ Consumers in wealthy nations are, in effect, outsourcing deforestation to the tropics.

Government Subsidies That Backfire

Governments sometimes accelerate deforestation through their own agricultural policies. In the Mexican state of Campeche, two subsidy programs for agricultural and cattle production, PROCAMPO and PROGAN, were found to have causally increased tropical deforestation. PROCAMPO increased deforestation by an estimated 15% in the state, while PROGAN increased municipal deforestation by about 7%.^{18 19} These subsidies were designed to support food production and rural livelihoods, not to clear forests, but the incentive structure made clearing profitable. Researchers have suggested that targeting livestock subsidies to areas with low deforestation risk and high existing productivity could maintain food output while limiting forest loss.

Urbanization and Population Growth

Cities themselves consume relatively little forest directly, but urban expansion triggers a chain of indirect effects. As cities grow into formerly agricultural land, farming gets pushed outward into forests. In Southeast Asia, urban expansion encroaches on non-natural habitats and increases disturbances to adjacent natural areas.²⁰ Globally, urban land area could nearly triple by 2030 compared to 2000 levels, with the highest growth rates projected in regions that were relatively undisturbed by development at the turn of the century, including biodiversity hotspots in East Africa, West Africa, and South Asia.²¹

Population growth shapes deforestation not mainly through urban sprawl but through increased demand for food, fuelwood, and farmland. In developing countries, population pressure drives both land fragmentation (dividing existing farmland into smaller plots) and land extensification (clearing new land). Migration plays a particularly underappreciated role: settlers moving to forest frontiers directly precede the most visible form of land conversion, the transformation of forest into farmland.²² Only a small fraction of migrants actually move to forest frontiers, which makes identifying who they are and why they go a potentially powerful leverage point for conservation.

Governance Gaps and Tenure Insecurity

Weak governance does not cause deforestation the way a chainsaw does, but it creates the conditions under which every other driver operates unchecked. In Nicaragua, decades of history in Indigenous territories show that problems in forest management are deeply linked to land-tenure insecurity. When communities lack clear, enforceable rights to their land, illegal logging and resource mismanagement flourish.²³

The flip side of this problem is revealing. Across the tropics, deforestation rates inside Indigenous lands are consistently lower than in non-protected areas, and they perform comparably to formally designated protected areas in reducing both deforestation and degradation.²⁴ In the Brazilian Amazon, Indigenous territories with full, formally recognized property rights show significantly less deforestation than those without such rights.²⁵ Loss of intact forest has also been considerably lower on Indigenous peoples’ lands than on other lands globally.²⁶ The evidence is fairly clear: recognizing and enforcing land rights for local and Indigenous communities is one of the most cost-effective strategies for keeping forests standing.

Do Zero-Deforestation Pledges Work?

Over the past decade, many large commodity traders and consumer-goods companies have signed “zero-deforestation commitments,” promising to purge deforestation from their supply chains. The results are mixed. In the Brazilian Amazon, the Soy Moratorium reduced direct soy-linked deforestation in participating municipalities by about 57%. But that amounted to only about 1.6% of total Amazon deforestation over the same period, because soy is just one commodity and the moratorium covered only one biome.²⁷

There is also a leakage problem. Traders with longer relationships to particular sourcing regions are more likely to adopt zero-deforestation commitments, but they also appear to have less effective ones, suggesting that the commitment itself may not change behavior on the ground as much as the pledge implies.²⁸ More broadly, corporate commitments have advanced the conversation around private-sector responsibility for forests, but implementation gaps remain large.²⁹ Voluntary pledges, in short, are a useful tool but not a solution on their own.

How Deforestation Changes Rainfall and Worsens Itself

Deforestation is not just a response to economic and demographic pressures. It also creates feedback loops that make the problem harder to reverse. Forests recycle enormous amounts of moisture into the atmosphere through their leaves. When large areas are cleared, the air above becomes hotter and drier, which reduces rainfall downwind and increases the severity and extent of drought. In the Amazon, accumulated forest loss since 2000 has warmed and dried the lower atmosphere enough to amplify both drought and fire.³⁰

Satellite data confirm that tropical forest loss causes statistically significant drops in rainfall at all scales measured. At larger scales, each percentage point of forest cover lost reduces annual precipitation by about 0.25 millimeters per month.³¹ Climate modeling since the late 1980s has consistently shown that replacing Amazon forest with pasture weakens the hydrological cycle: less evaporation, less rainfall, and higher surface temperatures.³² At a global scale, the climate feedback from deforestation actually reduces water runoff across most tropical regions, counteracting the increase that would otherwise come from removing the trees themselves.³³

The practical consequence is grim: deforestation dries out remaining forest, making it more fire-prone, which kills more trees, which dries the climate further. In a region like the Amazon, there is real concern about crossing a threshold beyond which large areas of forest can no longer sustain themselves and convert to degraded savanna. That process, once started, would be extraordinarily difficult to reverse.

Monitoring From Space

One thing that has changed dramatically over the past two decades is our ability to see deforestation as it happens. Satellite systems at multiple resolutions now allow governments, NGOs, researchers, and even private companies to track forest loss on roughly an annual basis. Combining different types of satellite imagery, from coarser systems that cover large areas frequently to fine-resolution sensors that can identify individual clearings, has made it possible to detect abrupt changes and identify their likely causes.³⁴ This transparency has been a prerequisite for enforcement, supply-chain monitoring, and public accountability. Without it, zero-deforestation pledges and government moratoriums would be unverifiable promises. Whether that transparency translates into action, though, depends entirely on the political and economic incentives in play on the ground.