Wednesday, December 16, 2015

Defense Indexes vs Mass Shooting Database >= 14 'Dead + Injured'

In my last post, I established that the FBI will complete the three year period (2013 - 2015) with ~64M NICS background checks for firearms purchase.  In this post, I look at the relationship between two Defense  index/exchanges traded over the same period. Specifically, I am interested in the whether there is a relationship  between high volume "spree shootings" and equity increases for defense companies. If widespread publicity of spree shootings encourages Americans to purchase extraordinary volumes of weapons, does this in turn equal greater purchase of defense stock? R code here.

Here are the eleven events with (Dead + Injured) >=  14 mass shooting incidents according to Mass Shooting Tracker. I have left out shooter names. Some of these events will be recognizable by location:

Monday, December 14, 2015

AR-15s, Shopping Carts and Gun Ownership in America

Update: December 2015 and 2015 the year broke records which rendered my 2.3M projection for December short. 3.3M NICS  checks were completed for December 2015 bring the yearly totally 23M. Both numbers are records. I will update my projected  December data at some point. - RMF

Last week's cover for the "New Yorker" shows a happy, clean cut, young couple picking up their AR-15 (?) with their milk and hand grenades in their shopping cart at a local super mart.  2.2M NICS FBI  firearm checks were completed in Novermber 2015. After the December totals come in, it looks like we will have over 23M NICS checks for 2015, an all time annual record. Ironically, the day of the San Bernardino shooting (12/2), the NYT announced we broke a record for single day guns sales on the last Black Friday. If we don't already, we probably now have more firearms in America than cars. This would make sense because more people now die from firearms than car accidents in America.  I made a quick projection of the 221M NICS gun checks in America for the last 17 years. Between 2013 - 2015, 29% of all gun checks for the last 17 years will have been processed. I projected December 2015 at 2.3M.:

 (21093273 + 20968547 + (19827376 + 2300000)) / (222363898 - 892840)
2013 + 2014 + projected(2015)  / (eighteen year total  - first incomplete year)
[1] 0.2898311

For the three years 2013 - 2015, there will be (projected) 64,189,196 (~64M) NICS FBI gun checks processed. If all of those checks were approved, that three year total would be equivalent to 20% of our 315M population.  More data on NICS Operations can be found here

This is a (table) matrix from the data found here.  December 2015 data projected. Click to enlarge tables and charts.

Friday, December 11, 2015

Gun Deaths in America: Suicides, Homicides, "Mass Shooting Tracker" Data


Important note: RMF:12/24/2015 Shooting Tracker data  is now maintained by Gun Violence Archive at

This is an attempt to parse some of the data from Shooting Tracker and talk about death by firearms in America. The project took me a bit longer than I wanted because that data needed normalization and corrections and because the subject is somewhat poorly researched due to Congressional limitations on research. See my R code here.  I bound together Shooting Tracker data from three years of  mass shootings which those publishers define as any shooting that injures and/or kills four or more combined. This definition gives us 1,347 dead and 3,817  injured  from "mass shootings" for the three year period 2013 - 2015 (to date).  Before we talk about this comparatively small number of deaths and injury, some background statistics are in order on gun violence in America. I assume the data is accurate and analyze it as I find it for all my sources including Mass Shooting Tracker, CDC and the FBI.

Tuesday, December 1, 2015

GIS based precinct maps: Ballots Good, Jack and Joy.

These are 2015 precinct election maps. GIS code in R is time consuming and complex. Click to enlarge these maps. The equal frequency (quartile) class intervals are expressed as: (min,max) for good ballots for the 2014 and 2015 general elections. The Jack Louws, Joy Gilifilen, and  precinct differences (Jack - Joy) totals for 2015 are expressed in color driven scales which have separate ranges (indexes) per map. Both precincts 182 and 183 have part of their data hidden in the totals by the state of WA to protect privacy. The shape files  and precinct maps can be found here. My R code (which is very derivative - see 1,2) can be found here (3,4).  If you are a friend and you need code and data, email me.