Gun deaths vs. gun laws

[BainsBane]

Last edited Tue Feb 26, 2013, 01:08 AM - Edit history (1)

Source: Occupy the NRA
https://www.facebook.com/OccupyTheNRA

[NYC_SKP]

1. Generally, it seems that stricter gun laws can be correlated with fewer deaths per capita.

However, many of the same correlations by state can be made with respect to other data points, i.e.: funding for education, level of education, property values, high school completion rate, etc., ad nauseum.

Which is to say, "OK, so what?"



PS: props to the graph maker for including Obama support as a criterion.

[BainsBane]

2. well, it shows a correlation

between progressive policies, including education and stricter gun laws, and lower gun death rates.

[NYC_SKP]

4. I'd argue that the causal relationship between stricter gun laws and fewer deaths cannot be made.

And that it's more about the broader social and political context, which happens to more likely include more carefully crafted gun regulations, than the regulations themselves.

[BainsBane]

6. the key factor is numbers of guns

more than laws per se. Laws may or may not be effective, but the more guns in a given population, the higher the number of deaths by guns.

[NYC_SKP]

8. I don't see "number of guns" listed as one of the data points.

Also, "Strength of gun laws" is a sort of sketchy and subjective entity, to me.

[BainsBane]

10. here is chart

As for your hypothesis about education, it is also possible that education (demographically) may influence likelihood to own a gun.

[krispos42]

17. Interesting.

Thanks, I'll save that chart.

Here's a couple of more, I don't know if you've seen them or not. I made them from DoJ data.


Ordinary least squares is a pain in the ass!

[BainsBane]

18. why multiple homicide rates?

And have you included suicides?

[krispos42]

19. I was curious to see if multiple victim homicides were affected by...

...the rise of the double-stack handgun in the 80's (13 to 17 rounds carried, instead of the usual 7-8) , and the semiautomatic magazine-fed rifle in the 90's (e.g., AR-15s and AK-47s with 30-round magazines).


The big drop in the homicide rate that started in about 1990 was driven by drops in single-victim homicides. The rate of doubles, triples, quadruples, and five-plus homicides has remained relatively constant.

No, I didn't include suicides, although I imagine the CDC has historical records.

[BainsBane]

25. Did you happen to see the Nova episode

"The Mind of a Rampage Killer"? They looked into efforts to figure out what prompts people like Lanza and Holmes to carry out mass murders. They advanced a few theories but said it's essentially impossible to predict who will behave that way, other than they show signs of anti-social behavior early in life (but then so do lots of people who never engage in rampage killings). They did say that propensity toward violence was easier to predict. They listed a number of factors, some of which were gender (male), age (between adolescence and late twenties), a fixation on violence, and ACCESS to weapons. That last one is key. Without access to a gun, they can't carry out their intent. For kids in poor, urban areas, access seems particularly easy since guns are evidently left around in public a lot, at least according to the accounts of some boys in a juvenile rehabilitation program in Wisconsin.

[krispos42]

33. Thom Hartmann calls testosterone something like...

..."the leading cause of murder throughout history".

I didn't see the NOVA episode, no... the only thing that I see that's up-do-date is "The Walking Dead", and occasionally "The Daily Show"!

[Recursion]

22. Huge drop-off in 1997 when the first post Roe v. Wade cohort turned 14

That, IMO, is the least coincidental thing about those graphs. That is, if a state has very strong or very weak gun laws and is not Vermont, I can predict with some confidence where that will fall on the gun deaths spectrum. But the signal itself is weaker than the variance band of the data, which is always a troubling sign.

[BainsBane]

27. a criminologist told me

The number one factor for violent crime is the percentage of young men in the population. Evidently there is a biological link between testosterone and violence.

[Jeff In Milwaukee]

29. I hear what you're saying....

Better social services and better education are key factors in reducing violence of all types, including both homicide and suicide by firearm. Strong gun laws play a role, but not the ONLY role and possibly not even the DEFINING role.

[BainsBane]

34. actually

He believes guns have no role, and that they bear little if any relation to violent crime or murder rates overall.

[Jeff In Milwaukee]

35. Oh dear...

Well that would be a pretty indefensible position, I think.

[BainsBane]

36. he should probably clarify

I might have made his view seem overly simplistic, but he is definitely strongly pro-gun and often argues against stats like these (not that the particular illustration in my OP is a shining example of transparency in data).

[krispos42]

9. And the non-gun-death rate?

Keep in mind that "gun death" includes suicides, which are the majority of gun-related deaths.



Of course, I would also be interested in knowing if the gun-control rating also gave points for useless, feel good laws like bans on assault weapons or 11+ magazines.

[BainsBane]

12. sure, of course that's an important point

I don't happen to have a chart of that. I'm sure we can access raw numbers by state through CDC.

[krispos42]

16. Generally speaking, there should be a correlation between gun ownership and gun-related death

Obviously, if the fewer households per capita have guns, there should be fewer gun-related deaths, all things being equal.


It is complicated, however, that states that have varying levels of social services, public schools, etc., that makes things not equal.

If a state is less inclined to treat domestic abuse seriously (good ol' boys club), then that state might have a higher rate of homicide overall, which would necessarily include a higher rate of gun-related homicide.

[Recursion]

23. The linear regression is a bad fit, though

The tails are easily-ordered but the median area is decidedly not.

[pintobean]

3. May I send you a friendly PM? nt

[BainsBane]

5. no, you may not

because there is no way it will be friendly.

[pintobean]

7. You're welcome to post it

if it isn't. I think it's important and it shouldn't be public. That's why I asked.

[BainsBane]

11. okay

but if it's rude, don't bother.

[pintobean]

32. You're welcome.

[JVS]

13. Keeping secrets from DU?

What shouldn't we know?

[Loudly]

14. Two observations:

Virginia rates a 44 on strength of gun laws? My impression of this state is that drive-thru gun and ammo stores are a tolerated business model.

~ and ~

Even the horrific gun crime rate in Chicago is diluted down by the peaceful burbs and rural communities.

[BainsBane]

15. the question of how they rate gun laws

obviously begs inquiry. I posted a chart in #10 that shows the correlation between numbers of guns and gun deaths.

[Recursion]

21. Yeah, that was my big question

Do you have the background on how they came up with the number?

[BainsBane]

24. No, I'm afraid not

This just came up as an image on my Facebook. See the chart from Mother Jones in post 10.

[BainsBane]

26. the chart lists the Daily Beast as a source

You could look there to see how they determined their ratings of gun laws.

[Recursion]

20. Interesting. I'm not convinced you can assign "strength of gun laws" a single number

It's a vector with several dimensions.

[BainsBane]

30. don't pretend you care about facts

at least be honest. If you want to examine the information behind the numbers above, go to the Daily Beast. I'm not interested in your Tea Party pro-gun propaganda. Guns kill. That's their design and their reason for existing. If you find that surprising, you're working hard to avoid thinking.

There is in this thread some discussion of how the categories are problematic and suggestions of other ways to approach the information. You could consider that if you were interesting in doing anything but spreading propaganda for the gun lobby. No amount of propaganda is going to change me into a sociopath with no regard for human life, so don't bother.

The CDC provides evidence of mortality rates. If you're going to try to pass yourself off as an empiricist, at least make a half decent effort.

[BainsBane]

31. 38,000 people die from guns every year

and another 73,000 are injured. Those are real human lives you purposefully dismiss.