In The Latest Misfires in Support of the More Guns, Less Crime Hypothesis Ayres and Donohue write:

In the wake of some of the criticisms that we have leveled against the Lott and Mustard thesis, John Lott appeared before a National Academy of Sciences panel examining the plausibility of the more guns, less crime thesis and presented them with a series of figures showing year-by-year estimates that appeared to show sharp and immediate declines in crime with adoption of concealed-carry laws. David Mustard even included these graphs in his initial comment on the Donohue paper in the Brookings book that PW refer to repeatedly in their current response. But Donohue privately showed Mustard as well as the Brookings editors that the graphs were the product of coding errors in creating the year-by-year dummies, and in the end Mustard conceded and withdrew them from his comment on Donohue.

Here is the graph and associated text that Mustard withdrew from his section in Evaluating Gun Policy:

Bogus graph showing crime decreases following carry laws A third criticism of the argument that the post-1992 years provide systematically different results is that Lott (2000) uses data through 1996 for his extensive array of empirical tests and arrives at qualitatively similar results as his earlier work. Furthermore, extending the data through 1998 produces results consistent with the basic Lott-Mustard conclusions. Figure 2 depicts the coefficient estimates from state-level regressions of the years before and after the Right-to-Carry laws go into effect for the four violent crimes.[3] Although before and after averages and averages of before and after trends can be useful because they provide some tests of statistical significance, they can sometimes be misleading. The year-by-year effects have the advantage of being more general than other models like the before and after time trends, the hybrid model and the two-year effects, which all impose more structure on the regression. The passage of the law is associated with sharp decreases in murder, rape and robbery. Murder and robbery rates are higher in Shall-Issue states prior to the passage of the law and fall immediately after the law goes into effect. As in Figure 1, these drops are much larger than would be warranted by a reversion to the mean explanation because the post-law rates are substantially lower than the pre-law rates. Rape shows a slightly different pattern. Right-to-Carry states have similar rape rates as other states in the years prior to the law. However, these states experience sharp drops in rape rates after the law is implemented. Aggravated assault rates in counties that pass Shall-Issue laws are higher both before and after the law goes into effect, with little change before and after the law.

[3] The control variables include all the demographic variables for age, gender and race, measures of per capita income, population density, measures of income and unemployment, state fixed effects and regional-year fixed effects. Regressions using county-level data produce similar results.