As I understand it cognitive biases aren't exceptions or adjustments to a model; they are observed phenomena. Any model of psychology or economics which purports to unify these phenomena must be able to explain/predict each bias individually. How does "attention economics" predict the planning fallacy?
Cognitive biases are exceptions or adjustments to the rational model: https://en.wikipedia.org/wiki/Rational_choice_theory On the whole, the field of Behavioral Economics is a correction to the Rational Economic Model. Behavioral Economics says "people are rational, except for the ways in which they are biased." They call it "bounded" rationality.
For instance, the planning fallacy is a correction to the idea that people will rationally predict how much time something will take. So we first estimate how long something might take, and then the planning fallacy teaches us to increase it to account for our bias.
> Any model of psychology or economics which purports to unify these phenomena must be able to explain/predict each bias individually
That's close to correct, but I'd like to distinguish explaining the bias vs. the data. The new theory should explain the data, not the biases in the old theory. Consider that Kepler's elliptical orbit theory didn't explain each individual planetary epicycle -- it didn't need to. Kepler's theory didn't need epicycles at all to explain the data.
Likewise, Attention Economics doesn't need a "Planning Fallacy", because it doesn't assume humans are good planners. It rather looks at how people actually allocate their attention while planning. Consider that if people allocate more attention to their plans, they are likely to make better estimates. So how are they allocating their attention when planning? In the "planning fallacy" [1], Kahneman and Tversky envisage "that planners focus on the most optimistic scenario for the task, rather than using their full experience of how much time similar tasks require." I haven't run the experiments myself, but one could certainly test for this in an Attention Economic experiment, by seeing how much more attracted people are to focus on the most optimistic scenario for their task, rather than the pessimistic scenarios. And then we can learn why they focus on the optimistic scenario, by manipulating other variables until we see which ones lead people to consider optimistic vs. pessimistic scenarios when planning.
Thanks for your detailed response. I'm afraid I'm still not sure how this simplifies much. The fact that people focus specifically on an optimistic scenario rather than having a random variance like a normal distribution seems very significant to me. If you can't predict this from first principles then you have to add in some extra explanatory factors, and then you still get your cursed epicycles.
