It is estimated that deforestation, forest degradation, and peat land emissions account for about 15 percent of total greenhouse gas emissions. A REDD (Reducing Emissions from Deforestation and Forest Degradation) agreement was reached at COP 16 in 2010; since then, REDD policies have been introduced with the goal of preserving forests and proposals have been put forward to compensate developing countries for avoided deforestation.
A recent paper in Food Policy studies the impacts of REDD policies on the agri-food sector and global food security. Such impacts are extremely relevant, as agricultural production and area expansion are primary drivers of tropical deforestation. REDD policies are likely to limit the expansion possibilities of agricultural land use and therefore may influence the competitiveness of the agricultural sector, agricultural prices, trade patterns, agricultural production, and ultimately food security.
The report assesses the impact of REDD policies on the agricultural sector using a global economic general equilibrium model (MAGNET) that analyzes the impact of these policies under varying scenarios. This model assumes that developing countries protect carbon-rich areas from deforestation and therefore lose the opportunity to use that land for agricultural production. In a series of simulation scenarios, the study protects an increasing number of area from deforestation and agricultural expansion. The associated impact on the agri-food sector (e.g. price, production and land use) and food security, both global and regional, is then assessed with the MAGNET model.
The various scenarios investigated include a baseline scenario, which represents a “business as usual” scenario based on conventional economic and demographic trends in the world economy for the 2010–2030 period and under an assumption of no new policy changes. The other nine REDD scenarios each protect a different amount of land/forests from agricultural expansion, ranging from the protection of 2 percent of potential agricultural land in the least restrictive scenario to 91 percent of potential agricultural land in the most restrictive scenario.
In the baseline scenario, the study finds that food consumption per capita increases in all regions; this indicates improving food security at the regional level. The highest per capita food consumption increase is found in Africa south of the Sahara (75 percent) and in India (44 percent); by contrast, OECD countries expect consumption growth of less than 10 percent. Overall agricultural land is estimated to expand by 7 percent, driven by rapidly growing demand for food products; expected increases in yields only partly mitigate this demand. The most pronounced agricultural area increases are observed in the less developed countries in Africa south of the Sahara (26 percent) and Southeast Asia (12 percent), where the highest population and income growth is expected during the study period. On average, 0.8 million hectares are predicted to be converted to agricultural land per year in the baseline scenario; this is nine times less than the 7.3 million hectares per year that were converted between 2000 and 2005. This large difference suggests that the majority of easily accessible and unprotected forest land suitable for agriculture has already been converted to agricultural land.
Examining the REDD scenarios, it appears that the implementation of REDD policies does not have a large impact on worldwide agricultural production. In the most restrictive REDD scenario (where 90 percent of potential agricultural areas are protected), only a 1.9 percent reduction in global agricultural production and a 1.6 percent reduction in global per capita food consumption is estimated compared to the baseline scenario. Similarly, total global agricultural area is expected to be around 8 percent lower in 2030, compared to the baseline scenario; the impact of this on food availability is largely mitigated by yield increases.
However, a regional analysis shows that the impact of REDD policies vary significantly by region. While highly developed regions are largely unaffected by REDD, less developed countries are significantly affected. For instance, in the most restrictive scenario, Central and South America, Southeast Asia, and Africa south of the Sahara decrease their agricultural production by between 12 percent and 18 percent. Similarly, per capita agri-food consumption in these regions falls by between 2.4 percent and 5.1 percent compared to the baseline scenario, and prices in these areas are also expected to increase significantly more (24 percent in Africa south of the Sahara) than the expected global increase of 7.6 percent. Based on these results, the paper argues that the overall impact of REDD on less developed countries, and on poor populations, depends on the restrictiveness of the REDD policies implemented. For instance, if 28 percent of the potential agricultural area is protected, then per capita food consumption decreases only by around 2 percent or less, compared to the baseline scenario, in the most affected regions.
The paper also highlights that the impact of REDD on agricultural sectors and food security in developing regions can be significantly mitigated, as has been proposed, through providing compensatory payments from developed regions to developing countries for avoided deforestation. These payments can be used to directly compensate for higher food prices or to stimulate advances in agricultural productivity. However, specific financing mechanisms to support this proposal have not been created on a large scale.
The paper argues that unless effective financing mechanisms are implemented, REDD policies should not try protect more than 15 percent of potential agricultural land, as any higher number would have significant impacts on the poor, especially in developing countries. Limiting REDD policies in this way would limit GDP losses and result in decreases in per capita food consumption of less than 1 percent, compared to the baseline scenario, and increases in regional food prices of less than 5 percent.