> Now you see rarely kids on the streets because they risk being hit by cars or kidnapped.

"Stranger" kidnappings are exceedingly rare. The majority of Amber Alerts occur due to custodial disputes between divorced parents when one parent takes the kid when they are not supposed to[1].

> Children (and parents) are often conditioned to be wary of strangers. However, in reality, only a small fraction of child abduction cases – around 0.1 percent – involve kidnappings by strangers or slight acquaintances.

[1] https://www.protection1.com/amber-alerts/

They’re less rare in certain communities. Something like 20% of abductions in the US are hispanic girls despite that demo being roughly 10% of the US population of children.

How many of those abductions are by strangers?

There is something to said about the risks of being run over especially given the popularity of SUVs which are so high up, but kidnapping is and has always been an incredibly rare crime for at least the past century. People are freaked out these days because of "Amber alerts" but these are nearly always cases of disputed custody between divorced/separated partners taking the kid against the wishes of the other parent, not strangers.

Conflating two wildly different things seems like it should be a formal fallacy. “I don’t go out without an umbrella because of the risk of it raining or an asteroid wiping out my town.” for example.

Car drivers kill thousands of people walking on the street every year. Kidnappers don’t.

Look at crime rates since the 2000s, or the 90s, or the medieval period.

This is the safest period in human history, more so if you live in a high income country.

Well, probably not, no. There's about 286 thousand years of anatomically modern humans running around pre-historically (unless you're meaning very literal history), and numerous illiterate societies which left us mundane archaeological records. Your conclusion is invariably cherry picked. Not to mention the goalposts for the concept of criminal have shifted drastically as has the means to enforce law.

Much of that 286 thousand years (or whatever) was spent in band-level hunter-gather societies. These societies have a number of nice features: People tend to be healthy, and in good times they may only work 20 hours per week.

But according to my anthropology professors, these societies tend to have very high death rates. For men, the lifetime risk of being murdered or killed in intergroup fighting runs about 10 to 20%.

So there's a real possibility that the modern era is very safe by historical standards.

>For men, the lifetime risk of being murdered or killed in intergroup fighting runs about 10 to 20%.

That makes being born in western Russia in 1905 look like a peaceful life

Your anthropology professors are ignoring the fact that they're inferring their rates from displaced populations.

Crimes of passion should also be exempted from the general murder rates.

Why should crimes of passion be exempted from the general murder rates? A society where it's more acceptable or common to kill someone who upsets you is a more dangerous society.

Because they're decontextualized. I'd posit that in the context of small organizations most of those murders are a product of deliberate and known risk. Given that, one could liken it to mutual combat at the ethical and social levels. That isn't the whole scope of the matter but the aggregate data would, I suspect, reflect a considerably different picture if it was investigated at such a resolution. Of particular interest would be infidelity and how that fits into the context of a given social order.

Safer in a statistical way, not an emotional / subjective / how does the news make us feel kind of way.

Or in other words, safer in a 'real' way as compared to an 'imaginary' way.

The idea of stranger danger is largely a myth, as most almost all crimes committed towards children are by those closest to them: parent, relatives and people in positions of trust.