I'd like to start giving people the option of commenting on shorter Cold Takes pieces (which I don't cross-post here or provide audio for). I'm going to use this post for that: I will generally leave a comment for each piece, and people can leave their comments as replies to that.
This was possibly my favourite email in the Cold Takes email newsletter so far. I always enjoy understanding someone's thought process before they've become an expert on a topic. Once someone knows enough, I think that their views usually change too slowly to easily see or demonstrate (one new piece of information or consideration, one new data point, naturally can't swing the holistic viewpoint quite so much when a person knows a huge amount).
It (unsurprisingly) reminded me of early Givewell material. Givewell is likely right more of the time now than in 2007. With more careful thought and knowledge built-up over time, comes better calibration. There is something lost though. How do we know that someone would change their mind in response to new evidence if we rarely see them change their minds? There is something wonderful about seeing people shift their views somewhat (or their confidence in their views) in response to transparent thinking in real-time. Anecdotally, this seems to happen a lot more in conversation than in writing (everywhere, not just in the EA community and adjacent spaces). In conversation, it is often much more acceptable to express uncertainty about conclusions while still presenting a framework for how you are thinking around an issue. It seems to happen rarely in public outside this community and adjacent ones.
My priors on the object-level question are very different to Holden's. My worst mental health happens when I feel stagnant/ can't contribute/am not valued. Being in physical pain with purpose has always felt much more bearable than having all the creature comforts of our modern time while feeling like what I spend my time doing is meaningless and doesn't add value to anything I really care about. This is obviously extremely weak evidence; memory is unreliable and I am a single individual. There might already be good evidence either way on whether feeling like your day-to-day life has purpose is a better predictor of subjective wellbeing than income or health for a sample size greater than one.
If hunter-gatherer societies consistently give everyone roles that visibly contribute to the lives of the people they know and love (eg. searching for food for the tribe) my prior is that this would feel more purposeful than modern day life (on average). If purpose is a more important factor in predicting subjective wellbeing than health or material wealth, then I would expect this study (across tribes/maybe even looking at biomarkers of depression instead of a survey) to replicate.