Gali and Gertler (1999) are the first to find that the baseline sticky price model fits the U.S. data well. I examine the robustness of their estimates along two dimensions. First, I show that their IV estimates are not robust to an alternative normalization of the moment condition being estimated. However, when using a Monte Carlo study to investigate small-sample properties, I show that the normalization chosen by Gali and Gertler (1999) yields a superior estimator. Second, I check whether or not the proportion of backward-looking firms augmenting the baseline model to fit the data is dependent on the type of contracting assumed. I find that using Taylor-style contracts, rather than Calvo-style contracts, this proportion jumps to 50 percent.