Firing Blanks? The following is analysis from my senior undergraduate thesis from 2015 (when I was an Anteater; Zot Zot) examining the effect of California's assault weapon's ban on crime rates. This paper uses a modified difference-in-difference estimator. The treatment group is Non-CA states, who let the federal assault weapons ban in 2004 lapse while California maintained a state assault weapons ban, serving as our control. The data examines per capita crime, various controls, as well as treatment/control and pre/post treatment for counties that lie exactly on the border of CA with Nevada, Oregon, and Arizona. I estimated the following regression: In(Murder)it = -8.36-0.02Povertyit - 0.16Treatit - 0.40 Postit (0.03) (0.5) (0.012) -0.06CountyUnemploymentit - 0.14(POST × Treat) (0.35) (5.57) (0.08) In(Murder) the logged murder rate per 10,000 people. Poverty is the percentage of the population within each county that is below the Federal poverty line. Treat is a binary variable = 1 for states that let the federal assault weapons ban expire (non-CA counties), O for CA. Post is a binary variable = 1 for years after the federal assault weapons ban expire. (2004 onward), or for pre-2004. County Unemployment is the unemployment rate. The interaction term, Post X Treat measures the interaction between counties within the treatment states. That is, the effect of being a Round to 2 decimal places and include a negative sign if necessary. A. Should I have run RTO instead? Assume at the 95% confidence level and a Z distribution. Answer "YES" or "NO". B. What is the T-statistic associated with the unemployment rate? Round to 2 decimal places and include a negative sign if necessary. C. What is the t-critical value associated with the 95% confidence level and a Z distribution? Round to 2 decimal places. D. What is the effect of being a treated state, before the treatment year? That is, what is the effect of being a county that removed an assault weapons ban before the year the ban expired. Include a negative value if the answer is negative, round to 2 decimal places. E. Is unemployment rate a determinant of the murder rate in this data? Answer "YES" or "NO" F. Is the treatment effect statistically significant at the 95% confidence level? That is, did removing the assault weapons ban have a statistically significant effect on the murder rate? Hint: Think about which coefficient we care about in a diff-in-diff set up. Answer "YES" or "NO". G. Does running this as a two-way fixed effects model make sense? Answer by inputting the number that is associated with the correct statement: