Yeah, I think the important tension comes out when there’s a project that seems very cost-effective on the object level but may have degrading effects on the procedural level (and those effects are hard to roll into a cost-effect model).
Do you have a hypothetical example? It’s hard for me to imagine such a case. Seems like most things that are cost-effective would have a (gross, if not net) strong positive procedural effect by making a healthier, more active, smarter demos.
Deworming programs are mildly coercive (strong social pressure to take a medicine that foreign NGOs think could be helpful) and normalize the practice of distributing state-approved drugs in schools (which I’d guess is procedurally bad), but we think they’re worth it because there’s a small chance of a huge object-level benefit.
Drug misinformation campaigns (like D.A.R.E.) propagate a bunch of inaccurate information about drugs through public school programs. Advocates think this is worth it because it scares kids away from drugs, but it has the bad procedural effects of degrading epistemics, spreading inaccurate information about human physiology, increasing the extent to which public schools are a venue for propaganda.
Yes, you are right.
I think it’s less interesting to EAs because we already buy the view that we should try to do cost-effectiveness comparisons of these things.
Yeah, I think the important tension comes out when there’s a project that seems very cost-effective on the object level but may have degrading effects on the procedural level (and those effects are hard to roll into a cost-effect model).
Do you have a hypothetical example? It’s hard for me to imagine such a case. Seems like most things that are cost-effective would have a (gross, if not net) strong positive procedural effect by making a healthier, more active, smarter demos.
Quick examples:
Deworming programs are mildly coercive (strong social pressure to take a medicine that foreign NGOs think could be helpful) and normalize the practice of distributing state-approved drugs in schools (which I’d guess is procedurally bad), but we think they’re worth it because there’s a small chance of a huge object-level benefit.
Drug misinformation campaigns (like D.A.R.E.) propagate a bunch of inaccurate information about drugs through public school programs. Advocates think this is worth it because it scares kids away from drugs, but it has the bad procedural effects of degrading epistemics, spreading inaccurate information about human physiology, increasing the extent to which public schools are a venue for propaganda.
Thanks, this is helpful. Not sure I agree that (1) has an overall anti-democratic effect procedurally, but I see the worry.