We are grateful to authors Carlisle Moody, John Lott, and Thomas Marvell (hereafter MLM) for their close attention to our article “The Impact of Right-to- Carry Laws and the NRC Report: Lessons for the Empirical Evaluation of Law and Policy,” which was published in the American Law and Economics Review (Aneja, Donohue, and Zhang 2011), and then re-issued as a National Bureau of Economic

Research working paper with some substantively unimportant errors corrected (Aneja, Donohue, and Zhang 2012). (Henceforth, we too will use the abbreviation ADZ to refer to our jointly authored work.) We think the attention to this work is warranted because it represents the most comprehensive and exhaustive analysis to date in the scholarly debate over right-to-carry (RTC) laws.

In thirty-three tables and twenty-seven graphs, our 2012 NBER paper showed that estimates of the impact of RTC laws were sensitive to four econometric modeling features: (1) choices of econometric controls (below we show results with variables used in work by Lott and David Mustard versus our own preferred models that include, most importantly, controls for incarceration, as well as models that do or do not control for pre-existing state crime trends); (2) functional form (estimating either an average post-passage effect or a change in the trend in crime); (3) choice of county or state crime data; and (4) computations of the standard errors (which, we showed, must include a cluster adjustment to reflect the lack of independence in state crime observations over time). The importance of these modeling choices is captured in Tables 1 and 2 below, which use data that is both more current and accurate than the data that was available to the National Research Council (NRC) for its 2005 report Firearms and Violence: A Critical Review.