In markets where the only imperfection is an environmental externality, a Pigouvian tax equal to marginal damages attains the first-best allocation. However, environmental externalities typically co-exist with other distortions in real world markets. In the presence of imperfect competition, a Pigouvian tax needs to be trimmed down because market power is already depressing quantities part of the way towards the first-best (Buchanan, 1969). In the presence of imperfect information between regulators and polluters, price-based incentives may be suboptimal altogether and dominated by quantity-based mandates (Weitzman, 1974).
In “Efficiency and Redistribution in Environmental Policy: An Equilibrium Analysis of Agricultural Supply Chains”, I tackle the question of how to design environmental policy in an industry where Pigouvian taxation is not feasible at the externality’s source, and where pre-existing distortions show up in the form of market power. The empirical context is the South American agricultural sector, a global commodity powerhouse with a major contribution to greenhouse gas emissions, primarily due to deforestation. This setting has two key features that shape the feasibility of environmental policy as well as its abatement performance. First, agricultural supply chains are funnel-shaped. The upstream stage consists of millions of atomistic farmers facing a concentrated sector of downstream agribusiness firms. Second, carbon footprints of agricultural commodities are mostly driven by the deforestation and production decisions of farmers. Hence, the externality’s source is upstream and varies geographically with the carbon intensity of land. These features suggest that while regulatory compliance is more feasibly ensured at the funnel’s bottleneck for a few large firms, downstream interventions can be poorly targeted because the incentives they ultimately need to correct are those of upstream farmers.
To evaluate the trade-off between policy feasibility and targeting, I develop an equilibrium model of agricultural commodity trade where upstream atomistic farmers face downstream monopsonistic firms, thus giving the supply chain its funnel-shape. I estimate the model using granular South American data, where the key empirical object is the supply elasticity of farmers and its heterogeneity across space. The main empirical finding is that supply is less elastic for farmers in geographically remote areas, resulting in wider markdowns on the farm-gate prices they receive. These remote areas also happen to be where the carbon density of land is highest, resulting in a positive spatial correlation between market power and the externality’s intensity.
Motivated by recent policy discussions on carbon border adjustments, I use the model to evaluate how a carbon tax on downstream agribusiness firms is transmitted to upstream farmers. Market power results in incomplete pass-through of the tax to farmers, and least complete of all to regions where markdowns are widest, which is also where carbon-intensities are highest. As a result, the Pigouvian signal, as measured by the pass-through rate, is eroded most in locations where the social cost is highest. Hence, the pre-existing distortion imperfect competition—worsens targeting for the policy tool that aims to correct the primary distortion, the environmental externality.
To the extent that market power unambiguously depresses output in a laissez-faire equilibrium, there is a classic intuition that “the monopolist is the conservationist’s friend” (Solow, 1974). Hence, my results introduce nuance to this view by showing that the presence of market power is more ambiguous for the transmission of environmental policy. Moreover, the sufficient statistic that resolves this ambiguity is the correlation between the two distortions—the degree of market power and the environmental externality’s intensity. In my empirical context market power worsens targeting because the correlation is positive. In other settings the correlation could be negative, in which case market power improves targeting by differentially raising Pigouvian pass-through where social cost is highest. Understanding how specific industries vary in the primitives shaping such correlations can help reveal general insights about the efficiency of corrective policies.
To conclude, the issue of tropical deforestation is a longstanding one and has attracted the interest of researchers in the social sciences for decades. Recent improvements in data quality (e.g., satellite imagery) combined with modern applied econometrics and equilibrium modeling have helped usher in a new wave of research seeking to understand the causes and consequences of deforestation, the role of market incentives relative to political economy, as well as optimal regulation (Souza-Rodrigues, 2019; Hsiao, 2021; Assuncao, McMillan, Murphy and Souza Rodrigues, 2023; Balboni, Berman, Burgess and Olken, 2023). At the same time, there has been a burgeoning literature on monopsony power in factor markets, part of which focuses on agriculture and the division of surplus between large buyers and small farmers (Bergquist and Dinerstein, 2020; Chatterjee, 2023; Rubens, 2023). This paper bridges these two literatures by seeking to understand the role of a particular pre-existing distortion—monopsony power—for regulating greenhouse gas emissions in agricultural markets.