I don’t think that’s the reasoning. You’ll note that the firearm ownership also starts well above zero while deaths per capita does not. Even just a pure numbers version (not per-capita, which is how you’re supposed to do any visualization where the population increases over time) would be dramatically flattened by proper inclusion and would not give the impression that they’ve ramped up disproportionately. Surely you can see that yourself, right?
I do data analytic work and visualization for a living, and all three graphs are very deceptive.
I’ll note that the second one is deceptive in the opposite direction, as firearm deaths are held to a lower axis rather than a higher one, auto deaths are well over an order of magnitude greater than firearm ones.
So all of them are kind of crap, but not consistently in the same direction.
