This was great! I'm a psychologist working in a neighboring area and there was still stuff here I hadn't heard before, like the aspiration treadmill.

One thing I wonder about: is fallible memory really a problem for the kinds of questions we usually want to ask about happiness? For the cold pressor task and colonoscopy studies where people prefer the experience where they technically experience more pain, we could say that people were "tricked" and they actually *should not* prefer that experience because there's more area under the pain-curve. But if they remember it for the rest of their lives as feeling nicer than the supposedly better experience, isn't that what matters? If we want to know what experience ultimately leaves people better off, we're mainly concerned with their remembering selves rather than their experiencing selves.

Fallible memories do seem like problems if we want to answer questions like "are people happier on Mondays or Saturdays?" and we only ask people retrospectively rather than using experience sampling methods (https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1559-1816.2008.00353.x). But usually what we want to know is "are these people generally happy right now?" and for that it seems fine if they've changed their minds or forgotten about how their past experiences felt in the moment. In fact, there's evidence that the negativity of negative experiences fades in people's memory faster than the positivity of positive experiences, and this helps people feel generally good in the moment (https://www.sciencedirect.com/science/article/abs/pii/B9780128000526000032). In that case, a fallible memory is a feature rather than a bug.

Anyway, thanks for writing––I'm looking forward to reading what you write next!

Expand full comment

Hey man, I was delighted to find this post and subscribe. My doctoral dissertation in progress is an argument that measuring subjective-wellbeing is a solution to a lot of the classical antimonies of welfare economics, and measurement issues form a big part of it, so this post was very useful: (See my thesis notes here if you're interested: https://philosophybear.substack.com/p/draft-zero-of-my-phd-thesis )

Anyway, I think for me the interesting tension is this- how good does a method have to be before it is, to use a phrase "good enough for government work" when it comes to measuring economic progress and the effects of policies? I think philosophers and methodologists like myself really love picking away at methodologies for the study of subjective-wellbeing, but what we often forget is that the other side of the equation for measuring welfare isn't some ultra sophisticated lovey-dovey holistic measure of social and economic effects, it's unweighted willingness to pay and cost-benefit analysis.

Indeed I sometimes think that methodologies like Kaldor-Hicks crowd out measures that I'd prefer *because they wear their weaknesses on their sleeve*. Whereas SWB measures as a yardstick for policy are *debatably* bad, Kaldor-Hicks is just *openly* bad- but something about that brazenness seems to win out.

How does one draw the line when it comes to measures of SWB. What sort of objections should be seen as intellectual curiosities, matters of concerns for academics, and what sort of objections should be seen as grave and direct enough to threaten a methodology's work in public decision making?

Expand full comment