A theory: the number of firearms (which translates into how accessible they are) isn't a linear scaling factor in the problem.
Comparative population could be relevant, although also doesn't explain the statistics you reference.
To try to prove/disprove the argument: has the prevalence (or nature) of bullying changed?
