There was a great study published in the journal Cognition a few years ago revisiting the famous kids & marshmallows self-control study, but where the researche

There was a great study published in the journal Cognition a few years ago revisiting the famous kids & marshmallows self-control study, but where the researchers additionally manipulated the children's beliefs in the environmental variability. They offered the kids some desirable thing and then either gave it to them or not, then checked how much self-control they exercised on the marshmallow task. The idea here is that if the payoff is highly unreliable, the rational thing to do is to go ahead and eat the first marshmallow because 1) there's a pretty good chance there aren't more coming and 2) even the first marshmallow might disappear. I think there's a pretty clear socioeconomic analog of this— if you have few resources and an unreliable environment, it's rational to act on a much shorter timeline.