The results of loosening gun restrictions

[bordelais7]

Murders rates are down in 3 of the 5 years since 2007 (using 2007 as the base).

I guess the bad guys legally obtained guns in 2007 when they may not have been able to otherwise, shot up their enemies, then put the guns away.

One Brow's response: look at the delta

based on this data:

2003 5.05 (5.7)
2004 6.15 (5.5)
2005 6.93 (5.6)
2006 6.30 (5.7)
2007 6.55 (5.6)
2008 7.66 (5.4)
2009 6.46 (5.0)
2010 7.00 (4.8)
2011 6.09 (4.7)
2012 6.46 (4.7)

The difference in the MO rate vs. the US rate becomes

2003 -0.65
2004 0.65
2005 1.33
2006 0.60
2007 0.95
2008 2.26
2009 1.46
2010 2.20
2011 1.39
2012 1.76

So, relatively sustained difference post-2007, although appears to be at least slightly trending towards catching back up to the national average.

Those numbers are probably far more mathematically precise than they deserve to be, but since I don't know the exact statistical significance I can't say that for certain.

 

[One Brow]

You going to make us a pie when you're done picking all them cherries?

Why did Missouri murder rates increase by ~30% from 2003 to 2007 while the national trend was to hold steady?

So, you are starting from 2003, and then accusing me of cherry-picking?

https://www.disastercenter.com/crime/mocrimn.htm

2000 6.2
2001 6.6
2002 5.8
2003 5.1
2004 6.1
2005 6.9
2006 6.3

Murder rates went from an abnormal low in 2003 back to the normal levels in 2004.

Please explain using only relevant legislation which was passed, or using the factors of change in policing, incarceration, burglaries, unemployment, or poverty.

Since the answer is statistical variance and the regression to the mean, none of those factors apply.

While you are responding, I will be working on the lattice for my pie. Maybe we can swap recipes after this ironically fruitless exercise.

You posted data supporting the position you were arguing against, and then after choosing what was an ideal year for you to form a counter-argument, accused me of cherry-picking. I agree, that was pointless.

Further, based on previous posts, instead of just acknowledging any of this, you'll probably accusing me of twisting things around and distorting what you said, although you won't bother to post it. That will also be pointless.

p.s., I'm (conveniently?) leaving for the next couple of days, so feel free to simply not respond and we'll call you the victor. I'm fine with that.

There's nothing to win. If the evidence supported your position, I'f have been happy to revise what I said.

One Brow's response: look at the delta

based on this data:

2000 6.2 (5.5)
2001 6.6 (5.6)
2002 5.8 (5.6)
2003 5.05 (5.7)
2004 6.15 (5.5)
2005 6.93 (5.6)
2006 6.30 (5.7)
2007 6.55 (5.6)
2008 7.66 (5.4)
2009 6.46 (5.0)
2010 7.00 (4.8)
2011 6.09 (4.7)
2012 6.46 (4.7)

The difference in the MO rate vs. the US rate becomes

2000 0.7
2001 1.0
2002 0.2
2003 -0.65
2004 0.65
2005 1.33
2006 0.60
2007 0.95
2008 2.26
2009 1.46
2010 2.20
2011 1.39
2012 1.76

So, relatively sustained difference post-2007, although appears to be at least slightly trending towards catching back up to the national average.

Supplemented with data from https://www.disastercenter.com/crime/uscrime.htm.

It will be interesting to see what 2013 add to the data.

 

[One Brow]

It seemed that using an absolute number (the increase of 60 homicides) was not as meaningful as using a rate or some relative number. Bordy's chart (homicide rate per 100,000 population) is more meaningful to me.

The press release gives an increase of roughly 16%.

 

[One Brow]

Murders rates are down in 3 of the 5 years since 2007 (using 2007 as the base).

I guess the bad guys legally obtained guns in 2007 when they may not have been able to otherwise, shot up their enemies, then put the guns away.

There were only two years after 2005 that were higher than 2005. Would you say they put away their guns after 2005, or that is an effect of choosing the highest year prior to 2007. 2007 was also an above-average year for the period of 2000-2007.


Just for reference, according to this gun shop, the law took effect Aug. 28, 2007.

https://www.topgunss.com/procedures.htm

 

[LogGrad98]

Maybe I am confused but I am not seeing how the per 100k numbers from a time significantly before and then after the legislation is not valid. I also do not see how that data supports the findings of the study. According to the normalized data, the rates have statistically been flat over the past 8 years, with a spike in 2008 followed by a decrease. I'm not sure why I shouldn't believe that data as opposed to an absolute rate of increase that does not normalize for relative population density. If I normalize for every other factor imaginable, but my population increases by let's just say 30%, I would expect a more or less 30% increase in whatever the variable is in the base number (or 20 or 5 or whatever the increase or decrease may be), but if I normalize for potential population changes then the resulting variable shows a more meaningful result. Pulling the number on a per 100k basis is a very easy, valid, and meaningful way to normalize that data. Bottom line, more people = more stuff happening.

The mean of the data for the per 100k numbers is overall 6.47, if you exclude 2003 (which would potentially be an outlier anyway), then the mean is 6.62, with a stand dev of .5, so none of the numbers are more than 2 standard deviations from the mean, and the trendline is nearly flat. So without breaking out minitab or anything there is really no meaningful statistical change no matter how you slice it, unless there is another way to normalize the data that shows a more striking increase over the time frame in question. From 2008 to 2012 on a per 100k basis there is statistically no change like the rest of the data, at least not from those few data points provided as the annualized numbers. If fact if you simply trend the numbers from 2008 on the overall trend is down.

The other thing is that in that press release they do not say what was used to commit those extra murders. It does not appear to claim an increase in gun-related homicides only, and it does not point that out. It would be interesting to see how the rates change by type of weapon used or method used. They do mention at the beginning of another paragraph that "the increase in murders with firearms in Missouri", but where they mentioned the increased number they made no mention of firearms. So they won't put a hard number to firearm deaths, but will allude to it later on. Standard spin technique.

It looks more like a case of spin the outcome to me. Look at this quote:

"There is strong evidence to support the idea that the repeal of Missouri's handgun purchaser licensing law contributed to dozens of additional murders in Missouri each year since the law was changed."

If they had statistically meaningful data that proved the assertion they wouldn't soften the statement with words like "support" and "contributed".

I would be interested in seeing the data behind the study and their actual findings instead of propaganda.

 

[One Brow]

Maybe I am confused but I am not seeing how the per 100k numbers from a time significantly before and then after the legislation is not valid. I also do not see how that data supports the findings of the study. According to the normalized data, the rates have statistically been flat over the past 8 years, with a spike in 2008 followed by a decrease.

Nationwide, there was just a decrease, with no spike in 2008. So, this description is pretty much in-line with what the press summary says about the paper.

I'm not sure why I shouldn't believe that data as opposed to an absolute rate of increase that does not normalize for relative population density. If I normalize for every other factor imaginable, but my population increases by let's just say 30%, I would expect a more or less 30% increase in whatever the variable is in the base number (or 20 or 5 or whatever the increase or decrease may be), but if I normalize for potential population changes then the resulting variable shows a more meaningful result. Pulling the number on a per 100k basis is a very easy, valid, and meaningful way to normalize that data. Bottom line, more people = more stuff happening.

So, you would prefer to talk about the increase as being per 100,000.

The mean of the data for the per 100k numbers is overall 6.47, if you exclude 2003 (which would potentially be an outlier anyway), then the mean is 6.62, with a stand dev of .5, so none of the numbers are more than 2 standard deviations from the mean, and the trendline is nearly flat. So without breaking out minitab or anything there is really no meaningful statistical change no matter how you slice it, unless there is another way to normalize the data that shows a more striking increase over the time frame in question. From 2008 to 2012 on a per 100k basis there is statistically no change like the rest of the data, at least not from those few data points provided as the annualized numbers. If fact if you simply trend the numbers from 2008 on the overall trend is down..

You don't just throw out outliers, unless you have a good reason to think they are wrong.

Nationwide, the trend is also down. The authors are not, to my knowledge, claiming an increasing effect over time, just a one-time shift that sees no counter-shift.

The other thing is that in that press release they do not say what was used to commit those extra murders. .

It's a press release about a paper. If the paper does not justify the summary, the authors will lose a lot of credibility. In particular, if the summary claims firearm deaths, and then the paper only looks at overall deaths, the authors will be exposing themselves to significant criticism for no good reason.

If they had statistically meaningful data that proved the assertion they wouldn't soften the statement with words like "support" and "contributed"..

Those are exactly the words I would expect in a scientific paper, particulary in the social sciences. Bold statements are rare.

I would be interested in seeing the data behind the study and their actual findings instead of propaganda.

I agree, it will be interesting.

 

[JAZZGASM]

https://www.breitbart.com/Big-Gover...rop-after-gun-control-laws-are-struck-down-2/
https://www.justfacts.com/guncontrol.asp

https://hotair.com/archives/2013/08...ncludes-gun-bans-dont-reduce-the-murder-rate/

 

[One Brow]

https://www.justfacts.com/guncontrol.asp

Just one example of the quality of this link:

40+% of households own guns.
0.5% of households used a gun in the previous year because "almost certainly would have been killed" if they "had not used a gun for protection."
There is no evidence at all the homes without guns have a lower murder rate than homes in the same neighborhood with guns.

Something does not add up.

 

[VINYLONE]

OB, since I don't really know your stance on firearms? So, could you please tell me? Are you anti-gun?

 

[fishonjazz]

For the record im not anti gun or pro gun.

I think that there should be restrictions on who can have a gun, which there are.

I have a 9 millimeter and a 22

My brothers and dad have lots of guns and hunt everything.

I have not shot my guns in years and if i didnt have em my life wouldnt change a bit

 

[One Brow]

OB, since I don't really know your stance on firearms? So, could you please tell me? Are you anti-gun?

I don't know enough about guns to have a solid, detailed opinion on them. My closest experience to them is when one of my mother's friends was shot outside our back yard. I've never owned a gun, and never felt the need to own one. Frankly, if someone is so fearful that they think they need a gun for self-defense, I'm more worried about the owner than the intruder. I have seen no convincing arguments that gun owners are safer from crimes.

On the other hand, many people on this forum describe that they go to some length to keep their guns safe, to keep themselves properly trained, etc. They are quick to condemn owners who handle guns sloppily or store them carelessly. I find it hard to object to people who follow these procedures owning guns. I have no objection to hunting with guns, training with guns, etc.

My current line of thinking is that gun ownership should be more of a privilege that you earn than a right you have to lose, but it's not set in stone.

The author of the original post has some recommendations to reduce gun deaths:
1) No gun sales to people who have two or more convictions involving alcohol or other illegal drugs
2) Minimum age of 21
3) No gun sales to felons

These seem like reasonable restrictions to me.

 

[LogGrad98]

Somehow they need to get at the mental health of individuals that commit gun crimes. It seems more often than not they talk about the treatment the person had been undergoing, often including psychoactive drugs. Yet almost none of the restrictions recommended for legislation get at the real 800 lb gorilla in the room. It is all about not giving guns to criminals, which is a bit of a "duh", but what about those with a history of mental illness but no real criminal past who snap and go into a school firing a 9 MM at everything that moves. I would have to look it up, but it seems it is rarely the case that a recidivist felon is responsible for shooting up a school. Most often it seems it is the quiet kid that never showed any signs of aggression, who by the way was on 4 different anti-psychotic and anti-depression meds for the past 8 years. When will that become part of the equation?