Impact on US Gasoline Prices of Eliminating Biofuels Production: An Equilibrium Analysis
18 Pages Posted: 1 Feb 2012 Last revised: 21 Mar 2015
Date Written: January 30, 2012
A combination of mandates, credits, and subsidies are employed in the US to encourage biofuel production. Recent demands for reduced federal spending have increased scrutiny of these incentives just as many are set to expire if not soon renewed by Congress. In this context, it becomes important to understand what impacts ethanol production has had on food and fuel markets and what changes can be expected if ethanol production is eliminated. We report here on a modeling study of these questions. Using a computable general equilibrium model that captures the market effects of the various interactions among fuel production, imports, exports, and consumption in equilibrium, we study two policy scenarios to elucidate the effects of ethanol production: (1) a future hypothetical scenario in which ethanol production is eliminated at the end of 2011 and (2) a historical counterfactual scenario in which ethanol production was eliminated in 1999, before it became a substantial proportion of US fuel markets. We evaluate these relative to a baseline scenario in which corn ethanol production peaks at 15 Bg/yr (in 2012) and stays fixed.
We also examine the sensitivity of model outputs along two parameter dimensions: the price elasticity of the global oil supply and the elasticity of US consumer demand for gasoline. Within acceptable ranges for both parameters, we find that wholesale gasoline prices between 2000 to 2010 would have been 1.7-2.5% higher on average without the contribution of ethanol to the fuel supply. For 2012 viewed apart, this effect is substantially stronger, showing a 6-8.6% increase in the wholesale gasoline price in the no-ethanol scenario.
Further, due to the fact that simulation results in these scenarios generally show prices for crude oil and refined petroleum increasing faster than those for corn and corn-based ethanol, the price disparity continues to grow larger even after ethanol production peaks in 2012. In 2025, the difference between the wholesale gasoline price in the baseline and no-ethanol scenarios is 12.5%, with a sensitivity range from 7.0 to 23.6%.
We compare these results to a series of recent estimates of similar measures from Xiaodong Du and Dermot Hayes. We find that both their published results [Du and Hayes, 2009] and a recent unpublished extension of these results [Du and Hayes] imply a consistently much larger price effect of ethanol production (a 16% average increase between 2000 and 2010 relative to the mean and a 41% marginal increase in 2010). Regarding their recent unpublished extension [Du and Hayes], we discuss a number of compounding circumstances that could contribute in whole or in part to these very high estimates. While we cannot compare our results directly to either of the estimates, because of significant methodological differences, we conclude that their highly nonstandard treatment of fuel imports and the selection of elasticity parameters well below typical values is likely responsible for the high result. Although highly inelastic fuel supply conditions, such as might lead to large price spikes given sudden supply shocks in ethanol production, are certainly not impossible in the very short run, we note two points: (1) a return to the peak levels of gasoline imports of 2006 itself could make up for more than half the fuel lost from a complete elimination of fuel ethanol, and (2) given the size of the ethanol production market in the US and the import market for sugarcane based ethanol, we find it unlikely that a single disturbance, be it from drought or public policy, could provide as strong a shock as considered in Du and Hayes (i.e., a complete and immediate elimination of all fuel ethanol production).
Keywords: biofuels, uncertainty, computable general equilibrium
JEL Classification: C68, D58, Q42
Suggested Citation: Suggested Citation