Biofuel policies (blend mandate or tax credit) have impacts on food and energy prices, and on land-use. The magnitude of these effects depends on the market response to price, and thus on the agricultural supply curve, which, in turn, depends on the land availability (quantity and agronomic quality) and relative prices. To understand these relationships, we develop a theoretical framework with an explicit representation of land heterogeneity. The elasticity of the supply curve is shown to be non-constant, depending on land heterogeneity and the availability of land for agricultural expansion. This influences the welfare economics of biofuels policies, and the possible carbon leakage in land and fuel markets. We emphasize that the impacts of biofuel policies on welfare and land-use change depend strongly on the potential development of the agricultural sector in terms of expansion and intensification, and not only on its current size.► Welfare effects of biofuel policies depend on the agricultural supply elasticity. ► We characterize the agricultural supply as a function of land quality heterogeneity. ► Elastic (US-like) and inelastic (France-like) supply cases are examined. ► Welfare effects of biofuel policies are examined accounting for soil heterogeneity. ► The impacts of biofuel policies depend on the potential development of agriculture.
In the last decade, several countries have supported biofuel production and set targets in terms of their use (Sorda et al., 2010). There are a number of political reasons pushing governments to promote biofuels, the main ones being climate change mitigation, employment in the agricultural sector, and energy security (Charles et al., 2007). As biofuel production at a large scale is not profitable in a context of relatively low gasoline prices (apart from the Brazil case), governmental targets would not be achieved without external incentives, and the recent increment in production has been driven by public policies and economic incentives (Kretschmer et al., 2009 and Sorda et al., 2010). For example, in the United States, ethanol production is supported by strong tax credits (VEETC: Volumetric Ethanol Excise Tax Credit) as well as by production mandates (RFS: Renewable Fuel Standards). In the European Union, biofuel consumption is also mostly driven by blending mandates and tax exemption. Biofuel policies are strongly distortionary, and generate welfare effects (De Gorter and Just, 2009a, De Gorter and Just, 2009b and Böhringer et al., 2009). In particular, the increasing production of first generation biofuels from grain and oilseeds participates in the increment in food price, jeopardizing food security. Biofuel production also generates environmental externalities, such as green house gases emissions or biodiveristy losses (Fargione et al., 2008, Groom et al., 2008, Petersen, 2008 and Tilman et al., 2009). These negative effects are due to land use change on the one hand, and market effects on the other. The magnitude of these effects depends on the elasticities of agricultural supply, and thus on the extensive (land use change) and intensive (intensification of production) margins in the agricultural sector. Biofuel policies may result in carbon leakage in fuel and land markets. In this context, it is important to understand the interactions between market effects and agricultural land-use to assess the holistic effect of biofuel policies.
To assess the environmental effects of biofuel policies, Searchinger et al. (2008) consider response of market, and estimate new crop supply and demand using historical conversion patterns. However, land-use is not modeled directly, and agricultural land expansion is not endogenous. Two main, complementary approaches are used in the literature to investigate the relationship between agricultural markets and land use change: computable general equilibrium (CGE) models and mathematical land-use share models based on partial equilibrium. The main difference between these approaches lies in their degree of complexity, the former approach being based on detailed simulation models, while the latter is based on stylized analytical models.1 CGE models make it possible to assess the impacts of biofuels policies on land use in a general equilibrium, using land supply curves (Banse et al., 2008, Keeney and Hertel, 2009 and Kretschmer and Peterson, 2010). The CGE approach provides powerful tools to simulate policy shocks, and to assess their impact on trade equilibrium. However, these models often assume Constant Elasticities of Substitution and Constant Elasticities of Transformation, and the key drivers of computed phenomena, like land use change, are not always apparent. Simpler mathematical analyses, such as land-use share models, make it possible to understand the key elements of the impacts of biofuel policies on land-use change. Evidence from the empirical literature strongly supports the notion that private land-use decisions are determined by the financial returns to different land uses (i.e., the Ricardian rent), and land quality consistently explains the aggregate distribution of land-use (Stavins and Jaffe, 1990, Wu and Segerson, 1995 and Hardie and Parks, 1997). For example, high quality land is typically allocated to intensive agricultural uses such as row cropping, while low quality land is often put into forestry. Land-use shares in a given area will depend on the distribution of land quality within this area.2Feng and Babcock (2010) use such a land-use share model to assess qualitatively the marginal effects of biofuel policies on land-use change and intensification, around equilibrium. However, biofuels policies are likely to modify agricultural production and consumption more than marginally, and the results of a broader analysis will depend on supply elasticities away from equilibrium (which are likely to be non-constant). This difference matters when one focuses on the welfare effects of biofuel policies. De Gorter and Just (2009b) conclude their article on this point, emphasizing that the shape of the agricultural supply curve is influenced by available land for expansion, which modifies the supply elasticity, and then the deadweight costs of biofuels policies.
The present paper proposes a formal framework to examine how the agricultural soil quality heterogeneity of a country influences the welfare implications of biofuel policies and their effect on land-use change. Our analysis is in line with the welfare analysis of De Gorter and Just, 2009a and De Gorter and Just, 2009b, completed by accounting explicitly for soil heterogeneity and its influence on agricultural supply. For this purpose we build on the framework of Feng and Babcock (2010). By specifying the form of the soil quality distribution, we extend their analysis in two directions. Firstly, the proposed extension allows us to determine agricultural supply functions and land supply curves as function of the quality heterogeneity distribution. We build such functions accounting for agricultural land expansion and extensive margins, as well as for intensification and intensive margins (i.e., the increase of input use and yield in response to output price increase). We show that the land quality heterogeneity distribution influences the shape of the agricultural supply function, which is likely to be non-linear.3 Application of our approach to US and France data illustrates our analytical results and emphasizes the flexibility of the proposed approach. Secondly, the proposed extension allows us to examine the effect of biofuel policies when equilibrium is modified more than marginally. We discuss how the heterogeneity of land quality influences the analysis of welfare implications of tax credit (De Gorter and Just, 2009a and Feng and Babcock, 2010) and blend mandate (De Gorter and Just, 2009b and Feng and Babcock, 2010). As the elasticity of supply curve is not constant, deadweight costs of biofuel policies vary with the availability of additional land in quantity and quality. In particular, the effect of biofuel policies on both land and energy markets have to be assessed to determine if there are carbon leakages in these markets. Our main message is that the consequences of biofuels policies depend on both the global land endowment of the country under study and the position of the equilibrium on the non linear agricultural supply curve. The possibility to develop further the agricultural sector is thus more important than its current size.
The framework proposed here would be helpful for further research examining analytically the indirect land-use change impact of biofuel policies in a context of trade between countries, or world areas, with different land endowment.
We developed a theoretical analysis of the impact of soil quality heterogeneity on agricultural supply curve, and examined how it modifies the welfare analysis of biofuel policies. Extending the theoretical framework of Feng and Babcock (2010) by considering an explicit soil quality distribution allows us to analyze the effect of biofuel policies beyond marginal analysis of the equilibrium. This provides a more accurate quantitative estimation of the impact of biofuel policies on land-use change, and food and biofuels production and consumption. This also allows us to complete the welfare analysis of tax credit policies (De Gorter and Just, 2009a) and mandatory blending (De Gorter and Just, 2009b) by describing the role of soil heterogeneity. By shaping the agricultural supply curve, land endowment influences the welfare economics of biofuel tax credit and biofuel blend mandate. The main message is that the scarcer the land available for agricultural expansion, the less elastic the supply curve and the higher the transfer form consumers to producers. Moreover, for a given tax credit, the scarcer the land, the lower the “water” in the tax credit (because the policy is less effective), and the higher the deadweight costs of underconsumption and overproduction.
Of course, further research are needed, both for practical application and theoretical analysis. From a practical point of view, it would be interesting to get a assessment of land quality heterogeneity for different countries,20 and to examine how the framework can be applied when there are more agricultural products and market effects. From a theoretical point of view, future research will use the developed framework in a bilateral trade approach, in the spirit of Keeney and Hertel (2009), with two countries having different land endowment. The non linearity of supply curves may induce interesting results when the magnitude of biofuel consumption increases beyond marginal effect. Such development should account for trade to examine the induced environmental effects of biofuel production and trade, in particular on indirect land-use change.
