We analyze the properties of Synthetic Control (SC) and related estimators when pre?treatment fit is imperfect. In this framework, we show that these are generally biased if treatment assignment correlated with unobserved confounders, even number periods goes to infinity. Still, a demeaned version SC method can improve in terms bias variance relative difference?in?difference estimator. also der...

Pretreatment is among the most costly steps in the overall biological conversion of cellulosic biomass to fuels and chemicals. Pretreatment also has a profound effect on all other operations from choice and preparation of feedstock all the way to final product recovery and use of process residues. Yet, because cellulosic biomass is naturally resistant to breakdown to release the sugars containe...