You might be surprised to learn that CEAs of mental health interventions in the EA space (example) don’t count the value of preventing suicide and self-harm. But on a DALY basis, self-harm and suicide have roughly the same burden in total as depression as a whole, precisely because they’re so much worse (this is to say nothing of effects on income).
I think that mental health interventions, and especially direct suicide counselling, might be really underrated, simply because the research hasn’t been done in a lot of depth (that I’m aware of). Part of this is because it’s quite hard to accurately quantify the effects of interventions on long-term suicide rates (you need many years, and access to death records). However, there are a few good studies that I think might point to ways this could at least be estimated, and then we can agree on an appropriate evidence discount.
The theory of change gets quite convoluted. For example, people may report feeling suicidal or having suicidal thoughts, but this may have no real correlation with their attempts. Or we may find that people who report feeling less suicidal commit suicide less often, but it may not be true that reducing their feelings of suicidality actually has any effect on their long-term attempt rate.
You might be surprised to learn that CEAs of mental health interventions in the EA space (example) don’t count the value of preventing suicide and self-harm. But on a DALY basis, self-harm and suicide have roughly the same burden in total as depression as a whole, precisely because they’re so much worse (this is to say nothing of effects on income).
I think that mental health interventions, and especially direct suicide counselling, might be really underrated, simply because the research hasn’t been done in a lot of depth (that I’m aware of). Part of this is because it’s quite hard to accurately quantify the effects of interventions on long-term suicide rates (you need many years, and access to death records). However, there are a few good studies that I think might point to ways this could at least be estimated, and then we can agree on an appropriate evidence discount.
The theory of change gets quite convoluted. For example, people may report feeling suicidal or having suicidal thoughts, but this may have no real correlation with their attempts. Or we may find that people who report feeling less suicidal commit suicide less often, but it may not be true that reducing their feelings of suicidality actually has any effect on their long-term attempt rate.
Here’s a good paper looking at 84,000 people’s PHQ-9 scores and follow-ups with suicidality. Unfortunately, the evidence isn’t causal, but this study looks at suicide prevention RCTs and finds that, at least, depression interventions reduce ideation, but not suicidality. I’ll be doing a bit more research in this area because I think it’s quite promising!