Founders Pledge report: psychedelics & subjective well-being

Founders Pledge is a nonprofit that supports entrepreneurs who have pledged to donate a portion of their wealth to high-impact causes. They recently published a research report recommending the Usona Institute. Here's their blog post about the report (a). Here's the full report (a).

This is good news! Many wealthy entrepreneurs look to Founders Pledge to inform their philanthropic giving, so this report could forerun a substantial inflow of donor dollars into the space.

Founders Pledge is also affiliated with the Effective Altruism community – the Open Philanthropy Project is one of its primary funders. We've put a lot of energy into making the case for psychedelics to EA audiences, and this has often been met with skepticism (Exhibit A). The Founders Pledge report is the first time an EA-affiliated organization has formally recognized the effectiveness of psychedelic interventions, which is something of a milestone, and great to see.

Psychedelics & subjective well-being

The Founders Pledge report can serve as a jumping-off point into many interesting topics, but let's zero in on just one for now – whether or not psychedelics can boost subjective well-being.

The FP researchers considered two basic benefits that psychedelic interventions might have: 1. Their ability to treat pathologized mental health conditions (e.g. depression) 2. Their ability to improve subjective well-being

Here's an excerpt about subjective well-being from the FP report:

The evidence on increased prosocial attitudes and behaviours and improvements in subjective well-being is weak. We found only one experimental study that used direct subjective well-being measures before and after taking a psychedelic, and it found no statistically significant improvement.[51] That said, two studies—a prospective[52] and an unpublished[53] one—found improvements on a composite well-being scale and multiple studies found self-reported, self-attributed improvements of subjective well-being, i.e. participants stated that they think the psychedelic experience improved their well-being and prosocial behaviour and attitudes.

I basically disagree with Founders Pledge here – from my perspective it seems that there is actually quite a lot of evidence suggesting that psychedelics can improve subjective well-being.

Let's dig into the study FP cites as finding "no statistically significant improvement" in direct subjective well-being – Griffiths et al. 2006.

Griffiths et al. 2006 is a landmark psychedelic study – it found that psilocybin could occasion mystical experiences in healthy people, and its follow-up (Griffiths et al. 2008) found that folks continued to view these experiences as intensely meaningful for a long time.

In their study, Griffiths et al. assessed subjective well-being in three ways:

  1. Direct measurement of affective balance using PANAS, administered before and after the psilocybin session
  2. Participant self-assessment of whether their moods & attitudes had been affected by the session
  3. Observations by other people who knew the participant – did they notice a change in the participant's attitudes and behavior after the session?

When Founders Pledge says that Griffiths et al. 2006 found no evidence of direct subjective well-being improvement, they're pointing to (1). And it's true – Griffiths et al. didn't find any change in PANAS scores that they could attribute to the psilocybin session:

None of the factors on the two widely used questionnaires assessing five factors of personality (NEO and PI-R) and measures of general positive and negative affect (PANAS-X) was differentially affected by psilocybin. At screening and at 2 months after session 1, there were no significant differences between the group that received psilocybin on the first session (N=15) and the group that received methylphenidate on the first session (N=15)

But Griffiths et al. did find that the psilocybin session improved participants' self-assessments, and community observers did notice statistically significant improvements in participants' behavior and attitudes – items (2) and (3). You can see all of these results in Table 4 of the paper.

So what's going on here? How is it that participants felt that psilocybin improved their mood & attitudes, and that other people in their life noticed positive improvements, but they still scored about the same when they took the PANAS after the psilocybin session?

One possibility is that the session deluded the participant – it caused them to think that things have improved (and caused people who knew the participant to think things have improved too?), but there was in fact no objective improvement in their affect when it was measured. The FP report doesn't get into it, but I suspect this is sorta what the Founders Pledge researchers believe.

Another possibility is that the psilocybin experience caused participants to reevaluate how they think about their lives, and this reevaluation messed with the psychometrics.

Here's what I mean: imagine that someone has an experience (say, a psychedelic trip), and this experience causes them to re-conceptualize the scale they use when they think about their well-being.

e.g. They used to assess themselves as an 8 out of 10 on some scale. Then they have this experience, and they think "Oh wow, now it seems that the scale I was using was wrong. I had been thinking that I was close to as good as I could be (8 out of 10), but now it feels like 'as good as I can be' is actually way higher than I thought..."

So they bump up the top end of the scale they're using. Now they assess themselves as 8 out of 50, not because they think their life has gotten worse, but because they're conceptualizing "as good as I can be" differently.

But this shift isn't reflected in the psychometric instrument they've been using, so when they assess themselves again, they have to normalize their new view to match the instrument's scale:

Something like this is probably going on when participants take the PANAS, have a psychedelic experience, and then take the PANAS again. This dynamic would explain why the "after" measurement didn't detect a substantial change from the "before" measurement – because experience causes the participants to relate to the scale differently.