I believe this is a result of cognitive mechanisms for maintaining consistency.
That’s not enough to explain it. We evolved hyperbolic discounting for a reason. Either the reason goes away as we age, or it’s not strong enough to counter this effect.
I suppose it could also be a spandrel in its entirety, but that would take more of an explanation than the idea that the change in hyperbolic discounting is a spandrel.
Honestly, I think Gwern’s evidence points to there being little or no inconsistent discounting, but the phenomenon seems so real to me that I find it difficult to take—I would rather ascribe it to an unknown flaw in the way they detected discount rates, than accept that people rally do have consistent values over time. Inconsistent values seem to match very well with the complaints many people have about their own behavior.
So, perhaps positing cognitive consistency mechanisms was a mistake. I do think we have such mechanisms, and I was entertaining hope that they might eventually win over temporal inconsistency… but the honest truth is that I think the cited study “must be wrong or limited somehow” rather than showing that adults overcome temporal inconsistency...
I’m talking about the relative weights we assign to good things at different moments changing as a function of how close those moments get. What distinction are you thinking of?
That’s not enough to explain it. We evolved hyperbolic discounting for a reason. Either the reason goes away as we age, or it’s not strong enough to counter this effect.
Careful. Spandrels exist (although it seems likely at this point that Gould overestimated how common they were).
I suppose it could also be a spandrel in its entirety, but that would take more of an explanation than the idea that the change in hyperbolic discounting is a spandrel.
Honestly, I think Gwern’s evidence points to there being little or no inconsistent discounting, but the phenomenon seems so real to me that I find it difficult to take—I would rather ascribe it to an unknown flaw in the way they detected discount rates, than accept that people rally do have consistent values over time. Inconsistent values seem to match very well with the complaints many people have about their own behavior.
So, perhaps positing cognitive consistency mechanisms was a mistake. I do think we have such mechanisms, and I was entertaining hope that they might eventually win over temporal inconsistency… but the honest truth is that I think the cited study “must be wrong or limited somehow” rather than showing that adults overcome temporal inconsistency...
hm.
“Temporal inconsistency” or “value inconsistency”? These are different.
I’m talking about the relative weights we assign to good things at different moments changing as a function of how close those moments get. What distinction are you thinking of?
When the changes follow different patterns for different things it starts getting difficult to infer values from choices.