Heroin now kills more people than guns:

[Boiler 45,519]

About 1.5% of private sales not requiring background checks fail. I do remember reading an article about shall we say people failing regular background checks, an offence, however these never seem to be followed through.

 

Criminals, well they have to have a history and I would imagine would be more likely to use a straw purchaser. Seems most have restraining orders etc, very few are what you might think of as serious histories.

[bcking 3,157]

1 hour ago, IDWAF said:

Ok, now we are on the same page.

 

Not saying that won’t make things better, because it would, except... the people against whom we WANT it to work (criminals) will still not comply with the rules.

 

As an example, many states claim that all private sales should have a BG check.  But I don’t want to pay for one when I sell a gun, and really don’t have to.  So I choose not to.  But I DO inject some due diligence by only selling to people I know well.  And even then, I have a bill of sale, DL, address, and have them sign a statement that they are not known felons nor are they in any way prohibited from buying a gun.  Is it bullet proof?  No.  A friend could lie.  

 

Now put me in a state that requires private sales BG checks... how does the state mandate it’s happening?  How do they know I have the gun, or am selling it?  Only HONEST people will follow the law.  But if I were a criminal, or selling to known criminals (because said guns command a higher price and I can make a profit, for example), I would just ignore this little tidbit of a law.

 

That’s my theory on why private sales checks will never work.  Just as I can sell prescription drugs on the street with no oversight, the same thing goes for guns.  Or just as I can fake a bill of sale to reduce the taxes I pay on a private sale car.   Always a way around it.  A criminal will always find a way around a law, I think you’ll agree.

The study I linked showed that states with universal background checks has lower firearm death rates, when adjusted for the presence of other gun control laws. 

 

If those two things aren't related, that means there is a confounder. Quite a massive one, given the RRs for those three are very significant and the confidence intervals are quite good.

 

Any ideas?

 

I can go back to the paper and pull out what they accounted for in their adjustments. It'll have to be tomorrow though. My wife doesn't like me on these forums

Edited March 15, 2018 by bcking

[Voice of Reason 10,298]

13 minutes ago, bcking said:

The study I linked showed that states with universal background checks has lower firearm death rates, when adjusted for the presence of other gun control laws. 

 

If those two things aren't related, that means there is a confounder. Quite a massive one, given the RRs for those three are very significant and the confidence intervals are quite good.

 

Any ideas?

 

I can go back to the paper and pull out what they accounted for in their adjustments. It'll have to be tomorrow though. My wife doesn't like me on these forums

 

I don’t have a subscription to your links above, but these are probably the states they are talking about, as well as their death rates per 100K:

 

CA - 7.9

CO - 14.3

CT - 4.6

DE - 11.0

MD - 11.9

NV - 16.8

NY - 4.4

OR - 11.9
PA - 12.0
RI - 4.1
WA - 9.0

D.C. - 13.8

 

Not sure how effective it is, looking at these numbers.  For reference the worst state was AK at 23.3, and best was MA at 3.4 (Thanks mostly to @spookyturtle)
Source: https://www.cdc.gov/nchs/pressroom/sosmap/firearm_mortality/firearm.htm

 

 

[bcking 3,157]

1 hour ago, IDWAF said:

 

I don’t have a subscription to your links above, but these are probably the states they are talking about, as well as their death rates per 100K:

 

CA - 7.9

CO - 14.3

CT - 4.6

DE - 11.0

MD - 11.9

NV - 16.8

NY - 4.4

OR - 11.9
PA - 12.0
RI - 4.1
WA - 9.0

D.C. - 13.8

 

Not sure how effective it is, looking at these numbers.  For reference the worst state was AK at 23.3, and best was MA at 3.4 (Thanks mostly to @spookyturtle)
Source: https://www.cdc.gov/nchs/pressroom/sosmap/firearm_mortality/firearm.htm

 

 

I can only get it when I'm on my office computer so I'll pull it back up tomorrow if I have time, and share more of their data.

[Voice of Reason 10,298]

31 minutes ago, bcking said:

I can only get it when I'm on my office computer so I'll pull it back up tomorrow if I have time, and share more of their data.

No sweat, neither of us is going anywhere anytime soon, and neither of our opinions will change the course of history.

 

Plus, I don’t want you getting into hot water with Momma over VJ.  I know how that can get...

[TBoneTX 45,292]

B., the PubMed links to those studies require TMC log-in; do you have open links for everybody, at least for the article Abstracts?

[bcking 3,157]

4 hours ago, TBoneTX said:

B., the PubMed links to those studies require TMC log-in; do you have open links for everybody, at least for the article Abstracts?

Oh sorry didn't realize you couldn't even see the abstract. I forget that when I copy the link it is including the TMC access.

https://www.ncbi.nlm.nih.gov/pubmed/26972843

 

Here is the study. I'll skip the systematic review for now. It just identifies 3 more studies that have similar findings (which shiws consistency, making the effect more likely).

 

It's the lancet though so there isn't free access, but at least this link will show the abstract. When I'm at work I'll pull out their methods and what they adjusted for.

 

[bcking 3,157]

Here is some more from the paper:

 

 

The crude results are the raw data. So that is the firearm mortality rates in the states with each law divided by the firearm mortality rates in the states without each law.

 

Crude values have some issues. First and foremost being states may have multiple laws and so the difference may not be due to that single law being studied. So that is when you look at the multivariable analysis.

 

In that analysis they basically control for all of the other laws. So that is essentially teasing out the relative risk that would be expected to be due to that single law. It also adjusts for population, overall homicide rates (so there is an adjustment for states that generally have higher homicide rates, guns and other means all included), unemployment rates, firearm ownership rates and firearm export rates.

 

I've already explained the Relative Risk (IRR) and the Confidence Intervals (CI), p value people more often understand (p <0.05 is generally considered significant, and that means that there is a less than 5% chance that the observed difference is due to random chance). The forest plot (the graph on the right) is a visual representation of the IRR and the CI (The square being each IRR, and the lines reaching out to the limits of the CI). The middle line at 1.00 would indicate there is no difference between states with that law and states without that law. If the CI line for any law crosses 1.00 generally you can say that there is no significant difference, though you could argue that more data may be able to narrow the confidence interval and yield a significant result in the future (Generally speaking the more numbers you have, the tighter your CI's are).

 

Obviously a limitation of the study is it is observational, and these are associations. Most people understand the idea that correlation doesn't mean causation and yes that is true. So initially if you were grading the evidence you would give it a low grade. However, you can upgrade observation studies when A. They have a large population, which this certainly does B. The association is strong (Which is true for a few of these laws, though not most) and C. We are unlikely to obtain higher quality of evidence. No one is going to set up a randomized controlled trial of states, it just doesn't work for this type of research. So observational is the best we are going to get. This is a very well designed observational study and is likely the highest grade of evidence we would get on the subject. They could do a similar study with more years of data (They used 2008-2010), which would be expected to narrow the confidence intervals. That may be helpful for some of the laws, but for the ones that fall clearly on one side or the other, it might create more certainty about just how much impact they have but it would be highly unlike to move the entire IRR to another position (Since, per the definition, there is a 95% chance that the true IRR lies somewhere within the confidence interval).

 

[TBoneTX 45,292]

The posted image doesn't appear (at least for me), but magnificent analysis -- and super description of RR, 95% CIs, etc.  Three thumbs up.

[bcking 3,157]

5 hours ago, TBoneTX said:

The posted image doesn't appear (at least for me), but magnificent analysis -- and super description of RR, 95% CIs, etc.  Three thumbs up.

Technology hasn't been my friend here.

 

Hope this attachment works. It's the forest plot from the study, their main results. Each law and how it impacts firearm mortality (The plot is of the multivariable analysis, so adjusting for every other law as well as the other factors I listed in my prior post).

 

More years of data would be expected to move the squares (RR) along each of their lines (CI), but wouldn't be expected to actually move them as a whole. If there is a potential confounder, something else that impacts firearm mortality that the model doesn't account for, that could move the square/lines together in either direction depending on what the confounder does.

[spookyturtle 9,465]

A close friend lost a family member from an overdose today. Trump is right about cracking down on dealers. 

[Dashinka 24,502]

2 hours ago, spookyturtle said:

A close friend lost a family member from an overdose today. Trump is right about cracking down on dealers. 

Terrible news.  Sorry for your friend’s loss.

[Voice of Reason 10,298]

2 hours ago, spookyturtle said:

A close friend lost a family member from an overdose today. Trump is right about cracking down on dealers. 

Sorry to hear that, buddy.  Always a grim day when this sort of tragedy strikes.

[spookyturtle 9,465]

21 minutes ago, Bill & Katya said:

Terrible news.  Sorry for your friend’s loss.

 

18 minutes ago, IDWAF said:

Sorry to hear that, buddy.  Always a grim day when this sort of tragedy strikes.

 

What’s even worse is a small child losing their father. It’s an epidemic that is out of control. And almost every death that I know about the person had been in rehab and clean for a while. A late 20’s guy I know has been clean for almost 3 years and he says he must always keep his guard up. It sucks them back in like nothing I have ever seen.