Chandio, R. and A. Katchova. “Sources of Bias in the USDA Baseline Projections.” Working Paper, 2023.
USDA’s annual Agricultural Baseline Projections contribute significantly to agricultural policy in the United States, hence their accuracy is vital. Although the bias in the baselines has been documented in the literature, its sources have not been evaluated yet. This study evaluates herding as a potential source of bias in the international baseline projections in two steps. First, we use a dynamic time warping algorithm to examine whether the projections for corn, soybean, and wheat variables are herded together, producing similar projection trends across different countries. Second, we compute the bias in projections and estimate whether the bias is higher across countries with more substantial evidence for herding behavior. We find strong evidence for herding of projection trends toward the United States. Our results highlight that herding is rational and error-reducing only for corn yield and wheat import projections, but not for other crops and variables. The findings of this study provide valuable insights for the USDA team producing the baseline projections and the stakeholders who use them for economic decisions. We contribute to the literature by proposing a new evaluation criterion of herding and potential adjustment framework for forecast accuracy optimization.
Chandio, R. “Conflicts and Conservation in the Amazon.” Working Paper, 2023.
This study evaluates how environmental policy restricting land use influences incentives for landowning farmers and landless peasants, and affects illegal occupations and land conflicts in the Brazilian Amazon. I use a ten-year panel data for conflicts, agriculture, and deforestation in the Legal Amazon region of Brazil and employ various matching methods to estimate the causal effect of an anti-deforestation policy on land conflicts and occupations. My estimates show the policy increases the number of illegal occupations in the Priority List municipalities while the number of land conflicts decrease. Analyzing the heterogeneity in the impact with respect to land values suggests that landowners and squatters both made strategic choices about whether to engage in conflict depending on the value of the land being contested. Landowners directed their efforts in capital and labor investments in their lands when the value of engaging in conflict to protect the occupied parcel of their lands was not more than the cost. But, for higher valued lands, landowners choose to defend their plots through violent conflicts.
Bevis, L. and R. Chandio. “Winners and Losers in India’s Green Revolution.” Working Paper, 2023.
The green revolution in India has been lauded for its positive impact on agricultural productivity and, in turn, the farmers’ lives. However, its effects on land redistribution and inequality, in the long run, are unclear. Large farmers may have benefited non-proportionately from the technologies that pervaded in the decades leading up to the turn of the 21st century. Richer farmers who own larger farms had early access to irrigation systems allowing them to profit more. Simultaneously, the smaller farmers who are unable to set up new irrigation technologies and thus cannot compete with the large farmers may be forced to sell their farms leading to systematic land redistribution. The poor farmers, in this process, may be worse off as a consequence of the green revolution in India. This raises the question of whether the policy implementation considered farmer welfare at all. Identifying the winners and losers in the green revolution is important not only to understand the current inequality in India but also to design appropriate agrarian and welfare policies. Moreover, this research has policy relevance specifically for countries in Africa that are more recent to new agricultural technologies and may inadvertently experience worsened inequality and welfare loss by non-optimal policies.