If people believed in statistics they would let their kids roam free like their parents and grandparents did. It's much safer than it used to be. Alas, US parents are overly protective.
I believe it's because parents and grandparents never looked at statistics and so because in communities where there wasn't a story of children getting kidnapped or being killed on the road then they just never thought about the chance happening.
It's far less safe in Mexico than in the US, though it's probably still safer in Mexico than Americans perceive it to be in the US.
> just somehow parents are more willing / less informed to take this risk.
You seem to presume Americans are well informed here; they are not. They consistently vastly overestimate the current risk and incorrectly perceive it as increasing when statistics does it is decreasing. Mexicans could be more willing to let their kids out because they are better informed (at least, have a more accurate perception of the risks) rather than less informed.
I don't have the stats you're looking for, but I would guess the prior probability of accidents and foul play in smaller communities is also going to be lower than it is for large urban communities, because these communities are more tightly-knit, and people drive less rashly to begin with, etc. Or maybe they drive more rashly - I don't know. The point is, there are a lot of factors to consider. So it's going to be hard to draw conclusions from stats.
In other words, how much of those numbers is explained by said lower probability, and how much by parents' willingness to trust their kids?
I wonder how much this has to do with the value of child life.
What are the statistics of Mayan children being run over / kidnapped?
I can't imagine it's somehow safer in Mexico, just somehow parents are more willing / less informed to take this risk.