I seem to have led our discussion into some clearly fallacious logical absolute terms.
If a person voluntarily chooses to engage in an activity, I take that as evidence they derive some value from it, because why else would they do it? This evidence can be undermined by a) direct evidence that the activity is harmful/not valuable, and b) reasons to believe the person would engage in the activity even if it had negative value, i.e. a cognitive bias motivating the activity.
So with drugs, we can see the health dangers, and pretty safely say that hyperbolic discounting will cause people to accept too many distant costs to justify the immediate benefits, and maybe also that the low-probablity high-magnitude harms aren't accounted for correctly.
Whereas with food, if people choose to eat it, I take that as a pretty strong sign that they enjoy eating it (for its price). If it's shown to be unhealthy, I might accept that hyperbolic discounting makes them undercount the downside.
Now, from the fact that a number of people write wikipedia articles without compensation, I conclude that a number of people derive value from writing wikipedia articles. I don't see any way in which this is harmful except the opportunity cost of their time. This cost occurs at the same time as or sooner than the benefit, so time discounting isn't an issue; it's a certain cost, so risk estimation isn't an issue; and I can't think of any other cognitive bias that would cause people to write articles if they didn't value doing so. It's only if we refuse to use people's choices as evidence that we have trouble trying to guess whether writing a wikipedia article is valuable.
Getting back to the original question, a lot of people choose to interact with financial institutions. Credit cards seem like a pretty clear case where hyperbolic discounting leads people into bad decisions. But a company like Goldman Sachs does most of its business with large corporations. These have a different set of biases than individuals, and I'm not sure which would cause them to overpay for underwriting or investment management services. There's the possibility that when Goldman and a corporate client interact they're somehow mutually benefiting at the expense of some third party, but my prior for that is low and I don't see much coherent evidence. |