Wednesday, January 02, 2013

Guns and Crime: U.K. Violent Crime.

I tried to access statistics for British crime from the national archive (http://www.nationalarchives.gov.uk/), but did not succeed (see Home Office).

The Daily Mail reports that Britain is the most violent EU country here: http://www.dailymail.co.uk/news/article-1196941/The-violent-country-Europe-Britain-worse-South-Africa-U-S.html. However, the numbers seem high compared to this Wikipedia graph (http://en.wikipedia.org/wiki/File:Violent-crime-rates-UK-1981-to-2007.png, from the Violent Crime article.).


The Wikipedia article Crime in the United States reports U.S. murder rate is well above that in U.K. and other EU countries (see Homocide). The article List of countries by intentional homicide rate shown rates for all countries. U.S. is high for developed nations.

Finally, not U.K either, but the Empires of Steel article Video Games and Real-World Violence has some interesting graphs for the U.S. I believe they were all taken from the U.S. Dept of Justice.

This post is mainly intended to collect data w/o much comment. Briefly UK trend seems very similar to US. Method of counting is different. If Daily Mail stats are correct, something very strange has happened recently. Wonder if the data is the same (same categories, same source, same counting methods) and what their source is.

Tuesday, January 01, 2013

Guns and Crime: U.S. Violent Crime Statistics

I posted initial gun and crime thoughts in my last post,  Guns and crime: Please stop misusing numbers. Now I'd like to talk about violent crime statistics.

Joyce Lee Malcom writes what seems to me a well-reasoned article: Joyce Lee Malcolm: Two Cautionary Tales of Gun Control. I think her numbers are good, but she perhaps fudges a bit to make her point. Australia and the U.K. are two countries not completely dissimilar to ours but with much tighter gun control. She says the numbers show it doesn't work. I'd like to look at some numbers for comparison.

First let's look at some raw number from the U.S. Department of Justice. I used the table builder to generate violent crime statistics for 1960 to 2010. I downloaded the spreadsheet (csv file) and made the following chart.


Murder and rape rates are small compared to aggravated assault. I separated them into their own charts so the trend would be clear.


To me two things stand out right away
  1. Crime rates are dropping since 1990. It's not like violence is out of control and needs a radical solution. It's getting better. We can argue about the best way to keep the trends going or improve them. But certainly we all should not be panicky, depressed and despairing about our country.
  2. Murder rates are insignificant compared to robbery and aggravated assault, and smaller even than rape. Yet murder gets all the focus, except when we focus on rape. Why? Well, probably because both those crimes are perceived as worse than robbery or assault. In 2010 the murder rate was 4.8, rape 27.2, robbery 119.1, and assault 252.3. So I guess we assume murder is over five times worse than rape and over 50 times worse than beating the crap out of someone.
In #2 above I'm being somewhat facetious. Nonetheless, the question of relative importance is a serious one that must be considered. Joyce Lee Malcolm cites a Brookings Institution conclusion that murders are down 3.2% in Australia since the National Firearms Agreement was passed in 1997. She also cites Australian Institute of Criminology statistics that show assaults up 40% since the 1990's. If that were to happen in the U.S. would it be worth it? The raw numbers for 2010 are 14,748 murders and 778,901 assaults. A 3.2% decrease in murders and 40% increase in assaults would change those numbers to 14,276 murders and 1,090,461 assaults. The differences are 472 less murders and 311,560 more assaults. Are 472 less deaths worth over 300,000 more people getting beaten horribly?

At this point you might argue that I'm reading too much into the numbers, that there must be other factors. You could be right. If so, I would argue the same about all the numbers presented in this debate. It requires thoughtful reasoned consideration. The problem needs to be properly engineered, not solved by emotional debate using suspect statistics.

There are many more numbers that I hoped to present but again my post has become too long. I hope this post at least gives convincing evidence that we are not in a crisis and that if we do need a solution it requires calm, reasoned analysis.




Guns and crime: Please Stop Misusing Numbers.

I'm not as smart as my friends. Some of them see the obvious that if we ban all guns violent crimes like what happened at Sandy Hook will go away forever. Others clearly see the obvious truth that if we criminalize guns only criminals will have guns. In my ignorance neither of these conclusions are obvious.

I'm also no lawyer. I do not understand all the intricacies of the second amendment. If the amendment is referring to self-defense against the government, I don't see how having a semi-automatic weapon helps me against a federal government armed with nuclear weapons and fighter aircraft. On the other hand, if it refers to self defense in general I don't understand why more women's groups are not against gun control. A handgun levels the playing field. Without superior weapons smaller and weaker people (of whom women are the lion's share) are relatively defenseless against larger and stronger people (usually men) with baseball bats.

Neither am I a doctor, psychologist, nor social worker. Certainly better treatment for those with mental health issues is desirable, but I have no idea how to best accomplish this. Hopefully informed debate on this will provide some solutions, however that doesn't seem to be the focus of any significant efforts currently under way.

What I do understand is numbers. While not a mathematician nor statistician by trade, I did study both as part of the course work for my B.A. in Math from UCLA. Furthermore my 30+ year career in systems and software often has involved mathematical problem analysis.

Almost everyone twists numbers to meet their needs. This is unfortunate. When it happens in the engineering world bad things happen. For example when a flight director ignores engineers' warnings about launching a space shuttle when the temperature is too low the shuttle might just explode and kill seven astronauts. Similarly, in a gun control or other debate, twisting numbers to meet your needs does not help in finding a solution. Unfortunately this misuse of statistics happens so often I feel many must have no idea what they are doing. I have a friend who is bad with numbers. However, he understands this and takes steps to compensate, including double checking his use of numbers with those of us who know better. If you likewise don't understand why most of the statistics trotted out in the gun debate are garbage, please stop using them or at least get advice from those who know better.

I started this post intending to bring in statistics I've found that I think speak to the issue. As usual, however, my preamble has stretched to the length of a full post. So I will conclude with the above appeal to be much more careful with the numbers you use and I will post statistics later.