Use the datafile BMN2016 on the textbook webpage. The authors report results for liquor sales. The data file contains the same information for beer and wine sales. For either beer or wine sales, estimate diff-in-diff models similar to (18.7) and (18.8) and interpret your results. Some relevant variables are id (state identification), year, unempw (unemployment rate). For beer the relevant variables are logbeer (log of beer sales), beeronsun (number of hours of allowed on-premise sales), beeroffsun (number of hours of allowed off-premise sales), beerOnOutflows, beerOffOutflows. For wine the variables have similar names.