Assault Weapons the Red Herring in the Discussion about Gun Control

 Data often has the annoying habit of being non partisan.  Sometimes it supports your point of view, sometimes it does not.  This is why it is useful to turn to it when you want to make legislation that will affect the lives of many.

I have been going over some data looking at murders, rates of murder, rates of murders that involved firearms, rates of murders that involved firearms per state, and so on.  In my previous post on this issue, I looked at absolute numbers of murders and the only correlations I could find between the numbers of murders related to firearms and total numbers:

1. When the absolute number of murders passes 200, most of these murders are committed with firearms.
2. Most murders in the US involve a firearm 
3. Most firearms used in firearm related murders are guns, not rifles, not assault weapons, just guns.

I really was not able to see other correlations.  I looked at Brady points (how gun friendly a state was) and could not really find any clear cut correlations there.  Reading another blog on a similar topic made me realize that I had ignored population size in looking at the data.  So I looked at the data (www.fbi.gov) again taking into account population size.

 In the graph above I ordered the states by what I am calling 'Brady points', this is the ranking the Brady Campaign (www.bradycampaign.org) gives states according to how strict their gun laws are.  The states with the highest points are at the left of the graph and those with lower points are at the right in order.  I am not sure I can say I see any kind of pattern.

There was a suggestion on another blog that perhaps southern regions with laxer gun laws  had higher rates of gun violence than northern states with stricter gun laws. So then I looked at the rates of both total murders and those related to firearms again ordering the states by region.  Again I am not sure that I can say I see much of a correlation, except that perhaps western states have slightly lower rates than the other regions, while southern states might have slightly higher rates on average (I will look at this more closely later).  

But then I also compared total murder rates with those related to firearms by region and on average it looks like western states have lower overall murder rates, and southern states slightly higher ones, but again will look at this a little more closely in the next few graphs.

The main point I want to make here is that just on this data, I don't see anything really jumping out for me, no nice simple trend.  So I averaged the data per region (I split the West into mountain and pacific as if you look at the Brady points, there is a very big difference between the two regions (west is the combined data for both mountain and pacific)).  So the main thing here is that the NE and the Pacific W, have much stricter gun laws than the other regions by quite a bit.  

One thing the data does show is the importance of controls, sometimes even the best of scientists forget to include them.  If you just look at the NE and the South then it looks like the area with the best gun laws has lower violence and lower gun violence.

 But that logic breaks down if you add 'controls', other states in this case, and )when you do, it is really hard to make that same correlation.  The only correlation I can see is that states with higher Brady Points (stricter gun laws) tend to have lower violence in general, but not necessarily lower gun violence.

So lessons learned?

1.  There is a correlation between guns and violence
2. The target of gun control should be guns, not assault weapons if we really want to cut down on the number of children being killed by weapons.
3. It is hard to find an easy correlation between gun laws and gun violence  (probably because most of them don't target guns)
4. There does not seem to be a correlation between region and gun violence (could probably find one if one looked hard enough)
5. Most people will not be killed by guns (look at the rates) and most people will not be murdered. 
6.  Data tends to be non partisan

I think the last point is perhaps the most important.  Emotions are very useful in the debate about violence and gun control as it is the emotions of people that typically spur them into action.  However, when that action leads to laws then the emotions should give way to unbiased data analysis.  By unbiased, I mean looking at what the data tells us and not what we want the data to say.  Parts of the data presented here is great for those who support gun rights, parts support stricter gun laws.  But the overall data is less partisan, but tells us that we need more information (more research) and fewer guns (not fewer assault rifles).  

I am in favor of stricter gun laws I always note this because I know that it is a bias that I have, and I think it is important for those who look at this data understand where I am coming from.   The way I look at it is that I live my life based on my biases (I don't posses a gun), however if I want to place my biases on others then I better look at the data objectively to see if it backs me up. 

My thoughts go to the families of all gun victims including those at Sandy Hook.