Thank you so much for writing this! I hope this isn't considered too off topic, but I run the Effective Altruism Addiction Recovery Group which I am maintaining but is still fairly slow at the moment. If you are reading this and are worried about your own addictive behaviors, feel free to join the server, or if you would rather not, feel free to reach out to me directly, and I would be happy to meet/help any way you think will be most useful. You should be able to join through this link:

https://discord.gg/W8sFnNEbdT

Pinea did complain about how many dimensions I wanted in my ethics...

Thanks, I see what you’re saying now. I can see value in positive reinforcement at least, but I guess I have a few reactions to some of the more specific points here:

1. Insofar as people can find reference classes they don’t fit that predict alcoholism, they can do the same for not drinking. Muslims, some other conservative theists, people with physical health conditions, people who are recovering alcoholics, people who rarely hang out with friends. I think you are at high risk if you are say a young atheist socialite in somewhere like NYC, and you can also count on very few of your friends being teetotalers. Given this I think the bigger difference the half non-drinker stat makes is to the risk of alcoholism if you do drink, which it doesn’t quite double, but probably something close to doubles (base rate I would guess is nearly a quarter of American who drink qualify as alcoholics at some point in their lives).

2. If there are sufficiently reliable things about someone’s situation/history that they can reference that it brings their risk down very significantly, then good! I would still probably disagree that it’s worth the personal risk, and think “drinking culture” is sufficiently bad that even if you fall in that bucket there are morally irresponsible ways to promote drinking, but if I can only convince people who do not have strongly mitigating factors not to drink (or heck, even just the people who have strong risk factors), then I think that would do a lot of good.

3. Honestly as I’ve mentioned here I don’t trust people that much to judge their own risk factors off of reasonable criteria. Too many people I run into spend literal decades in denial when they actually are already alcoholics by their own later admission. Heck, I have a close friend with a personal history of an addiction-like disorder, co-occurring mental health problems, and family history of alcoholism who I haven’t been able to convince not to use cocaine because they “don’t have an addictive personality”. I think the best heuristic in these cases is usually to just not assume you’re special and go with the averages, which don’t look great.

Sorry, I'm not sure I understand what you mean here:

"I meant for the stat of non-drinkers to be a positive signal for the general population to choose not to drink and still feel normie."

Could you rephrase? As for my stats, this is an example that's been helpful. I definitely agree that most people can eventually recover and stop drinking pretty much for good (or less reliably, in moderation). I'm currently sober for about two months, and hope to fully recover myself. What I meant is that even if you do eventually recover, there are huge costs that are typically incurred along the way in physical and mental health, personal relationships, career, finances, time, immoral actions, sense of identity/control, raw suffering, and certainly an above average risk of just dying before you recover (either from things like heart attacks and cirrhosis, or suicide). Even considering that most people can eventually recover, it seems like drinking isn't worth the risk of all this unless it was significantly rarer than it is.

Yeah, I’m getting the impression that one of the big things I ought to do with a final draft is expand my discussion of this change in my position, and possibly spin it off into its own appendix. For what it’s worth if this is true it means the risk from drinking is even higher than apparent, as even when you control for the portion of non-drinkers who are alcoholics or former alcoholics (depending on your preferred nomenclature), a quite significant portion of the people who don’t become alcoholics just don’t drink anyway (how much depends in large part on how many of these people used to drink a good deal and stopped, but never became alcoholics).

Thanks, this a decent gloss and I hope it will be helpful (I apologize again for the difficulty of the outline as currently written)

Worth adding though that alcoholism can get gradually worse over long periods, and many alcoholics spend decades in denial, so if you are trying to rule yourself into this class, you really should look at this much more objective criteria rather than sorta vibing "I've done this forever and I'm not an alcoholic".

This seems right to me.

I mean, people aren’t given “future alcoholic” cards. I think there are circumstances under which you can be sure drinking is especially risky, such as being a recovering alcoholic or having history with a different addiction or having a decent amount of recent family history with addiction, but I’m not aware of a ton of factors you can reference to be confident you won’t be one.

I don’t think your odds are more than half, but I do think they’re around one in ten if you’re an average American (if you’re drinking enough that cutting alcohol is a significant sacrifice, then I would guess your odds are worse than that). Those are very bad odds considering how badly this disorder fucks up your life and how small the upside is by comparison.

There’s also a risk of contributing to a drinking culture in which casual drinking is normalized or even expected in many contexts, hurting others who casually stumble into drinking because they didn’t realize how dangerous it was/just want to fit in (and drinking alone makes your odds even worse). I think we should treat alcohol a bit more like we tend to treat tobacco, so…the short answer is yes, I think almost no one should drink.

If anyone knows how to insert this table in my post, I would be very appreciative. I don't know if it's obvious or something, but I haven't seen any instructions for it and I am not technically skilled.