If you play Russian roulette and don’t die, it’s not proof playing Russian roulette was not a bad idea. Similarly, the fact that there was no nuclear exchange during the Cold War does not prove that the probability of it occurring was low. We know that it very nearly did happen, see Vasili Arkhipov. And people in that era gave high probabilities that it would happen.
So why didn’t they prefer to live in rural areas, or neutral or third world countries, places less likely to be hit? Many believed the media-promoted lie that the entire world would be destroyed by a nuclear exchange. But many didn’t.
For some, they could make the argument that to leave their friends and family for low wages in the middle of nowhere in America or a third-world malarial jungle is worse than a 1/3rd chance of death where they are. But they still should have been more eager to migrate if the right opportunity presented itself. There were many Americans willing to live in the third world, to work for wages higher than those in America, enjoy a lower cost of living, study abroad, or simply “to get away.” Yet it doesn’t appear that avoiding death in a nuclear holocaust was one of the motivators.
Perhaps they didn’t want to vocalize it. They don’t want to write to their mothers “hey, India’s great, my company’s paying me more, the culture is interesting, and if there is a nuclear exchange you’ll get fried and I won’t.” But some would have made the point less explicitly. Few of the sexpats are clear about their motivations, yet it became a common fodder for jokes. But there were no jokes about the guy holed up in the Philippines because he was convinced a nuclear war would occur. Why?
People are bad at judging abstract risks. The dangers of smoking, obesity, and crime are salient to people because they see them in their personal lives, but they subconsciously devalue the risks of nuclear war, placing it in the same part of the brain as the doomsday predictions of the local shaman which never got around to occurring. This will be relevant when hostile artificial intelligence becomes a real risk. Then, as during the Cold War, people will give high probabilities to the doomsday scenario. Yet many will not advocate the rational response of not building any AI.