Follow

An explanation I gave a while back as to why gun statistics generally favor the fact that less restrictive gun laws means fewer violence and homicides.

I basically explain why the typical argument of "countries with more guns have more violence" is inherently anti-scientific as it violates fundamental statistical analysis good practices. Instead we would use statistical causality tests for this, not correlation tests. When we actually look at the data from that perspective it generally shows that countries with less restrictive gun laws effectively lowers a nations homicide and violent crime rates.

Reattached the graphs from that post, but best to click the link to the original post where I go into more detail (the graphs are there as well).


QT: qoto.org/@freemo/1037666922745

๐ŸŽ“ Dr. Freemo :jpf: ๐Ÿ‡ณ๐Ÿ‡ฑ  
@hansw@mastodon.social Great I think that will giveus a foundation to work with. So Now ill provide some data, let me explain a bit how the grange...
ยท ยท 3 ยท 10 ยท 11

@freemo Your graph on England & Wales is misleading.

Counting methods were changed in 1998 to include additional offences, but your graph shows it as a direct comparison. In fact, violent crime peaked in 1995 and has declined steeply since the FIrearm Act.

Source: ons.gov.uk/peoplepopulationand

@mtorpey No not misleading, not entirely.

As your link points out the change in reporting practices took place in April, 2002. We are looking for spikes in the years/months following a change in gun laws.

The gun laws in the graph changed in 1989 and 1997. Meaning the change in reporting happened 13 years after the first change in gun laws and 5 years after the second. In both cases we notice a continued upward trend in violence following the gun laws, the second such increase being very significant and lasted the whole of the 5 years up until the change in reporting. After the change in reporting isn't even needed for consideration to prove that point, a 5 year trend is more than enough to establish granger causality in this case.

It should be noted that the 1987 act was very limited in scope, and as such the increase following it was not particularly dramatic, though does show a continued increase. However the 1997 act was much more broad and banned all handguns entirely, which is the primary mode of self defence, and as such shows a very significant effect even in the 5 years before a change in reporting methods.

@freemo I think we may be looking at different things. To quote my source above,

"The Home Office Counting Rules for police recorded crime were expanded in April 1998 to include certain additional summary offences. Figures before and after that date are not directly comparable."

Your graph shows a dramatic increase in crime around 1998, and another around 2003, both due to a change in counting rules and not due to an actual increase in crime. Your graph compares incomparable figures.

@mtorpey Ah yes I was looking elsewhere. Let me look into that and get back to you. It should be stated the fact that dozens of other countries show the trend listed here still strongly supports the conclusion. Thought that particular data point may well need to be removed.

@mtorpey so to take your consideration under advisement here is a different chart which does not have the issues with reporting you pointed out. In this case it represents homicide rates, not violent crime, and therefore not effected by the change in reporting practices of general crime. Furthermore the chart shows adjustments for anomalies in the data to provide a more accurate picture.

Again we see a significant upward trend right around the year of the handgun ban and with a consistent but more gradual upward trend that followed the earlier 1987 ban. The data appears to still hold true once the difference in reporting is factored out.

@freemo This seems to be more accurate. And we can indeed see an increase in the 3 years following the 1997 Act.

Not so sure about 1987 though, which shows only a very gradual increase after it.

In any case, you're right that this is just one data point, and we need to consider the trend across all countries rather than cherry-picking a few graphs of individual countries. Has a wider analysis been done on this? I'd love to see the result!

@mtorpey

Has a wider analysis been done on this? I’d love to see the result!

I have seen a few over the years that use this approach. Sadly due to the highly political nature of the subject good quality studies, or even studies that show consensus overall, are hard to find.

I am sharing a post I made from some time back so sadly I dont have the links on hand of the studies I had seen around the time I was researching this topic. Because the studies were lacking in this regard I also had to do my own number sleuthing to get to the bottom of it. While my personal analysis and conclusions obviously should be taken with a grain of salt not having been peer reviewed as a professional data scientist I obviously have some confidence in my methods when there is a lack of good peer reviewed articles. Typically even when I do my own investigation I like to use the body of peer reviewed articles to help me be aware of and avoid pitfalls (the reporting issue you mentioned being a good example of this of course), so that would have been helpful for me to have and disappointed in the lack thereof.

Anyway as I recall from my research at the time I was only able to find one tudy that actually looked at causation analysis, and it wasnt even peer reviewed.

@freemo Unfortunate, since this sort of research could save a lot of lives.

Similarly, I've always been disappointed there's no broad meta-study on how effective bike helmets are. Being in the Netherlands, I guess you'll see a lot of people not wearing helmets, and I've never seen a convincing source on how much safer (or less safe!) I'm actually making myself by wearing one.

@freemo I can care less if gun laws have an effect on crime rates. I don't care if an armed society has a higher homicide rate. I simply believe we have a right to own weapons for sport and to defend ourselves. I do not believe that the right to own weapons should be infringed upon. If I'm not jeopordizing the lives of innocent people, why shouldn't I be allowed to possess a MK 19 (fully automatic grenade launcher) ?
I'm not being sarcastic.

@Diptchip I have no issue with people owning fully automatic weapons. Generally they are not considered any more effective at mass killing, in fact they are considered less effective according to military training. Fully automatic is only ever used for random cover fire and never to hit a visible target.

@freemo I just think it's crazy that people try to justify having the right to own and carry weapons. I don't think statistics belong in any conversation regarding this right. You either have the right to defend yourself by any means necessary, or you don't. The latter is ridiculous to me.

Sign in to participate in the conversation
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.