Earlier quoted context omitted.
Or departments with more military gear are in more risky areas where the use of force is more likely. I could not read the article because it is blocked for me.
The study controls for that variable: > The newspaper’s analysis used the military’s database and paired it with a database of fatal police shootings from across the state, controlling for statistical variables like community income, rural-urban differences, racial makeup, and violent crime rates. The article explicitly cautions against drawing conclusions about causality. My intuition is that there’s truth to both e…
If it's both effects at the same time, in a feedback loop, it's actually ok - you can stop the loop by cutting either effect.