Agricultural Commodity Prices: A US-Brazil Perspective

 Between 2000 and 2015, land-use change, mostly caused by tropical deforestation (Mitchard, 2018), is thought to have caused 10 to 12 percent of all CO2 emissions (Le Quéré et al., 2016; Edenhofer et al., 2014). A lot of trees are being cut down because the world economy is growing quickly, energy prices are high, biofuels are getting subsidies, and the real prices of farm goods like grains have doubled (Mitchell, 2008; Alexandratos, 2008). Between 2000 and 2012, Kissinger et al. (2012) found that large-scale agriculture was responsible for about two-thirds of deforestation in Latin America and one-third in Africa and Asia. Again, about half of this deforestation is caused by growing foods for export to places like the EU, China, and North America (Lawson, 2014). As a result, countries like Brazil have promised to use a mix of market-based and command-and-control policies to cut down on deforestation. The question this paper tries to answer is whether these kinds of measures can stop trees from being cut down because of higher commodity prices. Our study looks at how well three main policy changes made in Brazil worked. Different strategies aim to stop different kinds of deforestation. The blacklisting of municipalities (PM) program targets towns and cities that cut down a lot of trees. It does this by tightening the rules for subsidized rural loans and doing more tracking and law enforcement.1. 

This strategy is mostly concerned with how much forest is being cut down at the municipality level. 

The industry-led Soy Moratorium (SM) aims to keep the commodity supply chain free of soybeans that come from land that has recently been cleared of trees. Thus, it focuses on how growing soy causes trees to be cut down. Some parts of land are regulated by conservation zones (CZs). We define "Conservation Zones" in this paper as three main types of protected areas: indigenous lands, sustainable use conservation zones, and highly protected conservation zones. We look at the edge of destruction in the Brazilian Legal Amazon. This part of the Amazon, which is still the world's biggest forest, is likely to have been hit by the most intense pressure to cut down trees so far. Our main set of data is a balanced panel of 470 towns from 2002 to 2013 that covers about 3 million km2. The main focus of the study is on cutting down trees outside of protected places. To start our study, we figure out how much direct effect agricultural product prices have on cutting down trees. We use foreign real prices to make a price index for each municipality. We use weights that are based on how much land each town farmed for each crop in 2002, which was the first year of our sample. Like Hargrave and Kis-Katos (2013) found, we also find that higher prices for farm goods are linked to more trees being cut down. We believe that if prices go up by 100%, there will be about 40% more trees cut down. The average 56% higher level of the price index from 2004 to 2013 compared to 2003 leads to an extra 1,700 km2 of trees being cut down every year. 

This is equal to about 19% of the 91,000 km2 of forest that was cut down in our sample over the ten years 2004–2013.

Next, we make an estimate of how the policies change the effect of the prices of international agricultural goods. This is the major contribution of the paper. We use a triple difference model (DDD) to look at how the prices in each municipality combine with policy exposure. This model basically looks at how prices change in places where a certain policy applies and how prices change in places where the policy does not apply. People in different cities and over time are exposed to a strategy in different ways. We can't say that there weren't common differences in deforestation before the policy was put in place. This suggests that our design effectively nets out possibly confounding factors that were driving both deforestation and the policy roll out. We discovered that putting towns on a "blacklist" cut the effect of commodity prices on deforestation by about 40%. This saved 35 km2 of forest per municipality every year. In the area we looked at, this result saved 9,000 km2 of forest. It makes sense that this would happen since the policy is supposed to make cutting down trees more expensive. Studies from the past have also suggested that this strategy cut down on tree cutting. We do not find a strong statistically significant effect for the farm commodity price index due to the soy moratorium. There are actually two effects that work against each other: the price of soy has less of an effect on deforestation under the Soy Moratorium, while the prices of other crops have a bigger effect. 

This fits with the idea that the soy moratorium will cut down on the loss. 

One example could be corn. Due to the soy moratorium, we see that deforestation is more affected by the price of corn. This may help explain why corn production has grown so much in the Brazilian Legal Amazon since 2006. We think that about 20% of the leakage to crops other than soy can be explained by leakage to corn. Studies of the Soy Moratorium that didn't look at other crops may have overestimated how much they affected forests, according to our results.  Lastly, we see that conservation zones make the impact of the prices of farm goods bigger. In general, prices were 40% higher in the years after zone increases than they were in the years before. This caused about 6,000 km2 more trees to be cut down outside of the conservation zones than would have happened if the zones hadn't been expanded. One way to look at our result is that the cutting down of trees might have continued into the new protected areas even if there wasn't a policy in place. The results are the same if we count the loss of trees in protected areas, where tree loss has been low in the past. Conservation zones take land away from the available land supply, which can make it easier for trees to be cut down on land that isn't protected. Based on explicit deforestation pressure and deforestation in places that aren't protected, our analysis shows that conservation zones haven't been as good at stopping deforestation as other studies have found.