In April, President Joe Biden announced that the United States will donate $500 million to the Amazon Fund, a program run by the Brazilian government to fund conservation projects and combat deforestation in the world’s largest rainforest. If world leaders hope to preserve the Amazon’s role as an important bulwark against climate change, more such international payments would be prudent—and could be a reasonably cost-effective way of reducing carbon emissions, according to research from the Climate Policy Initiative’s Juliano Assunção, Chicago Booth’s Lars Peter Hansen (a Nobel laureate), Argonne National Laboratory’s Todd Munson, and Columbia’s José Scheinkman.
The Amazon’s plant life soaks up 1.5 billion tons of carbon emissions every year, and contains a far greater quantity of stored carbon. The forest is so vast that it is most easily described on the scale of countries: it is only slightly smaller than the contiguous US, but shrinks by an area nearly equivalent to Kuwait every year due to deforestation. Much of this deforestation occurs in Brazil, which contains around 60 percent of the Amazon, and is done to free up land for agricultural use, especially cattle farming.
About 17 percent of the Brazilian Amazon has disappeared due to deforestation in the past 50 years. Some scientists estimate that once 20–25 percent of it is gone, it may reach a tipping point that will disrupt its hydrologic cycle—the process by which rainwater is recycled via vegetation and returned to the atmosphere—meaning large portions of it could turn from forest into savannah.
Because the Amazon’s carbon-capturing capacity benefits the whole globe, while the costs of forgoing increased agricultural activity in the area affect Brazil alone, it may be prudent for the international community to pay Brazil to limit deforestation. But how big should such transfers be, and what effect would they produce?
Assunção, Hansen, Munson, and Scheinkman explored these questions using a dynamic spatial model meant to capture the trade-off between using land for agriculture and preserving or restoring it as forest. The model incorporates the large degree of variation that exists between areas in the Amazon in terms of their productivity as agricultural land and their ability to absorb carbon.
“It’s not just a case of doing less deforestation; it’s also an issue of where you do it at,” Hansen says. Treating every square mile of the Amazon as equal in its capacity for productive agriculture or carbon capture would obscure the picture of what’s gained and lost by a particular allocation choice.
The model allowed the researchers to predict how land would optimally be allocated to maximize the welfare of Brazil given a particular price per ton of carbon emissions. For instance, looking at how land is currently allocated, they were able to calculate a “shadow price” of carbon that reflects the trade-off Brazil is currently willing to accept between agricultural productivity and carbon emissions. They find that price to be around $6 per ton—and at this price, with no international payments to change local incentives, the model predicts that the Amazon will cross the threshold of 25 percent deforestation within about 20 years, and that in the next 30 years, deforestation will produce net emissions of 31.6 billion tons. (Brazil’s total CO2 emissions for 2021 have been estimated at 2.42 billion tons.)
A relatively small transfer could make a big difference
Compensating Brazil for forgoing agricultural activity in the Amazon to stimulate reforestation could significantly alter the trajectory of land use in the rainforest compared with a business-as-usual scenario of no transfers.
The researchers note that their study is intended to be a benchmark for assessing alternative ad hoc policies. Nevertheless, they find that “the optimal strategy under the calculated shadow price would eventually produce enough deforestation to threaten the survival of the Amazon as a tropical forest.” However, the payments required to change the trajectory of deforestation in the Amazon may be far cheaper than the total cost to society of releasing that carbon into the atmosphere. If the international community paid Brazil $20 per net ton of carbon captured by the Amazon, the optimal policy for Brazil would be to restore enough forest to capture 16.1 billion net tons over that same three-decade timespan. By comparison, the US government has pegged the social cost of carbon at $51 per ton, and some estimates place the number several times higher.
As with many cost-benefit analyses, there is a justifiable concern about what Nobel Laureate Friedrich Hayek warned was a “pretense of knowledge” caused by overly ambitious economists seeking to influence policy outcomes. To counter this worry, Hansen and his collaborators quantified the uncertainty in the productivity inputs across the various sites in the Brazilian Amazon that go into their computations. Their analysis goes further by investigating how this uncertainty impacts the policy-relevant spatial dynamic implications of their model. Specifically, they find policy makers’ concerns about uncertainty should not lead to inaction but should instead suggest changes in the proposed course of action as it plays out over time and space.
Of course, implementing a socially efficient system of transfers would require the cooperation of not only the various members of the international community providing the payments but also the Brazilian government, whose attitude toward deforestation has shifted dramatically between various presidential administrations. Moreover, the researchers note that their analysis doesn’t take into account some potentially important costs Brazil may face from the erosion of the Amazon, including the loss of biodiversity and “the effect of deforestation on agricultural productivity in the Amazon or in other regions in Brazil, a country that is currently the fourth largest agricultural producer and exporter in the world.”