Percy writes:
A more meaningful correlation for comparing countries would be between the homicide rate and the gun ownership rate, where the number of guns owned isn't considered. The gun ownership rate could be by household or by individual.
That makes sense, I think. I'm not good at visualizing "rates" in my head... it takes me a while.
But this would imply:
IF gun-purchase-rate in the US went down (let's say... everyone stopped acquiring more guns for whatever reason)
THEN homicide rate would drop as well?
Even though the "amount of guns existing in the public" is a very high number.
That makes a certain amount of sense.
Like you're saying... the gun collector thing.
Those that collect guns, buy 1 every now and then and store them. Barely (perhaps even never) using them.
However, if there's trouble-a-foot (like, a gang war or large-scale intentions of violence) then there's a motivation to acquire "more and more and more" guns... and use them. This would drive the rate-of-gun-ownership up.
Is that the general idea?
That makes sense to me.
Here's a scatterplot I made and posted in Message 3018. It shows that by state there is no apparent relationship between gun prevalence and gun homicides:
Right.
This seems similar to the scatter plot I linked to in the article (#18).
As you're saying... number of total guns doesn't seem to make much difference... but the rate at which guns are being acquired does.
This graph shows how "total guns doesn't seem to make much difference..." but doesn't show anything about the rate at which guns are being acquired. I think?
Here's a great paper originally posted in Message 3151 that found that each percentage point increase in gun ownership resulted in a .9% increase in gun homicide:
This one I find confusing again.
Is this saying that the number of guns does make a difference? More guns = more homicide, regardless of rate-of-acquisition?
I suppose if the rate is increasing... and this results in more homicides... then during that time of rate-increase a graph that shows number-of-guns vs. number-of-homicides will show that more guns = more homicides.
We would need another graph of number of guns vs. homicides after the rate stopped increasing in order to see if it's just the rate or number-of-guns alone.
Perhaps that's my problem causing confusion? Just the timing at which certain data points are selected for viewing purposes?
Edited by Stile, : Wrong quote name