How to Halt and Reverse Deforestation
Our comprehensive new review provides evidence-based solutions
August 18, 2023. With record-shattering heat this summer, it’s a good time to consider how deforestation—the second-largest cause of climate change after fossil fuels—can be reversed.
Kalifi Ferretti-Gallon and I have just published a comprehensive review of what drives deforestation and how to stop it, in the Review of Environmental Economics and Policy. In our new meta-analysis, we systematically examined more than 15,000 data points from 320 published studies to see which factors are consistently associated with more deforestation, less deforestation, or neither.
What drives deforestation, and what stops it?
We found that deforestation is consistently higher near roads and cities, as well as where agriculture, livestock, and timber are more prevalent. Deforestation is also consistently associated with higher population and greater wealth.
Conversely, deforestation is consistently lower where policies directly influence allowable land-use activities (e.g. protected areas, Indigenous management, conservation payments, law enforcement, community forest management, certification of sustainable commodities), or where forests are passively protected due to steeper slope or higher elevation.
Deforestation is not consistently affected by policies and institutions that primarily seek other ends (e.g. democracy, general good governance, peace, or land-tenure security), nor by education or gender.
What have we learned?
Many of our findings aren’t too surprising. “Everybody knows” that agriculture drives deforestation, and that protected areas slow it down.* In these cases, the value of our study is providing quantitative, authoritative support for such statements.
But in some cases, our findings are surprising, or new, or go against frequently made claims. For example, people often suppose that poorer people are driven to deforestation to meet subsistence needs. While there are certainly poor people clearing forests, evidence shows again and again that there's more deforestation in places where people are richer. After all, richer people are more able to buy machines, access credit, or hire workers to clear trees.
Likewise, many hope that increasing agricultural yields would lead to growing more food on less land, reducing the need for deforestation. But this isn’t the case generally. Just as often, higher yields lead to higher profits, and thus higher incentive to deforest. Similarly, it would be convenient if democracy, peace, good governance, or secure land tenure, led to less deforestation. But in general, they don’t.
Here’s something interesting, that our meta-analysis is the first to find: Deforestation happens more when it’s hot. So there’s a vicious circle: deforestation is driving climate change — and it’s also fueled by it.
What’s new?
Although there have been previous reviews of drivers of deforestation, our study is the largest and most comprehensive of its kind. Our findings are broadly consistent to those of previous reviews, as the table below shows. But there are eight factors that we reviewed in our meta-analysis for the first time: agricultural yield, conflict, democracy, energy activity, gender, supply-chain initiatives, temperature, and trade openness.
One of the previous studies was our own 2017 meta-analysis, also published in REEP. Kalifi and I were motivated to update our earlier study for several reasons. Most importantly, in 2013 Matt Hansen and his colleagues published a groundbreaking global map of annual tree-cover loss. This revolutionary data set caused the number of studies of drivers of deforestation to skyrocket. Other remote sensing breakthroughs soon followed, allowing researchers to more easily study drivers of reforestation and forest degradation too.
What’s changed since our 2017 meta-analysis? Now, timber activity is associated with more deforestation. Community management is now associated with less deforestation, while wetness is no longer associated with less deforestation. Evidence is stronger that payments for ecosystem services (PES) reduce deforestation. The trickle of evidence that Indigenous peoples are associated with less deforestation has become a flood.
In our new study we didn’t just look at deforestation; we analyzed reforestation and forest degradation too. Reforestation is consistently higher in places with steeper slopes, consistently lower closer to cities and in places with higher population. Forest degradation is higher in places with higher populations.
Bringing it all together to halt and reverse deforestation
World leaders have committed to fight climate change by halting and reversing deforestation by 2030. In practice, the best example of a large-scale reduction in deforestation is from Brazil, which reduced Amazon deforestation by 80% between 2004-2012, even as agricultural production increased, as Frances Seymour and I described in our book, Why Forests? Why Now?
This achievement was the result of a concerted combination of many of the policies identified in our meta-analysis, including new protected areas, Indigenous territories, enforcement of forest protection laws on private land, credit restrictions for high-deforesting municipalities, and moratoria by the soy and beef industries on purchases from recently deforested land.
As leaders in Brazil, South America, and beyond look to halt and reverse deforestation, our new study can help guide policies and investment toward actions that support those goals, and away from those that don’t.
Busch, J., Ferretti-Gallon, K. (2023) “What drives deforestation, reforestation, and forest degradation? An updated meta-analysis” Review of Environmental Economics and Policy. doi.org/10.1086/725051
* A bit of clarification is in order. When I say that a factor, say protected areas, is consistently associated with lower deforestation, I don’t mean that protected areas ALWAYS lead lower deforestation. Rather, in the places and times studied by the papers in our data base, protected areas are correlated with lower deforestation more often than they were correlated with higher deforestation in a multiple regression analysis, and this difference is statistically significant.