Knittel and Smith present seven alternative versions of our model all of which apparently show a lower impact of ethanol production on gasoline prices. Four of these models are based on the flawed assumption that one can use the change in refiner’s margin and the change in gasoline prices interchangeably. The remaining three models all suffer from an obvious and endogeneity problem that when corrected results in results that are similar to ours. What is then left of their paper is a series of regressions of unrelated variables without appropriate controls, and with predictable results. I believe that the magnitude of all our results are reasonable and that they can be used in the current policy debate. Our results show that the closure of ethanol plants will have a serious impact on gasoline prices.
Hayes, Dermot J., "Second Response to Knittel and Smith" (2012). CARD Working Papers. 549.