This is a cross post written by Andy Masley, not me. I found it really interesting and wanted to see what EAs/rationalists thought of his arguments.
This post was inspired by similar posts by Tyler Cowen and Fergus McCullough. My argument is that while most drinkers are unlikely to be harmed by alcohol, alcohol is drastically harming so many people that we should denormalize alcohol and avoid funding the alcohol industry, and the best way to do that is to stop drinking.
This post is not meant to be an objective cost-benefit analysis of alcohol. I may be missing hard-to-measure benefits of alcohol for individuals and societies. My goal here is to highlight specific blindspots a lot of people have to the negative impacts of alcohol, which personally convinced me to stop drinking, but I do not want to imply that this is a fully objective analysis. It seems very hard to create a true cost-benefit analysis, so we each have to make decisions about alcohol given limited information.
I’ve never had problems with alcohol. It’s been a fun part of my life and my friends’ lives. I never expected to stop drinking or to write this post. Before I read more about it, I thought of alcohol like junk food: something fun that does not harm most people, but that a few people are moderately harmed by. I thought of alcoholism, like overeating junk food, as a problem of personal responsibility: it’s the addict’s job (along with their friends, family, and doctors) to fix it, rather than the job of everyday consumers. Now I think of alcohol more like tobacco: many people use it without harming themselves, but so many people are being drastically harmed by it (especially and disproportionately the most vulnerable people in society) that everyone has a responsibility to denormalize it.
You are not likely to be harmed by alcohol. The average drinker probably suffers few if any negative effects. My argument is about how our collective decision to drink affects other people. This post is not about what will happen to you if you continue to drink. It’s about what will happen to vulnerable people if you and I and others continue to drink.
Alcohol is a much bigger problem than you may think
Indian men’s alcohol consumption is the single strongest predictor of spousal violence. Regardless of wealth, education, employment or location, an Indian woman is much much more likely to be assaulted if her husband drinks. Indigenous Mexicans who drink every day are 13 times more likely beat their wives.
Mark Kleiman, a drug and criminal justice policy expert at New York University’s Marron Institute, argues that the research on the alcohol tax is very clear.
“The single most effective thing you can to reduce crime right away is to raise the price of alcohol,” he told me. “If you talk either about crime policy or drug policy, that’s got to be the number 1 recommendation — just because it’s so easy. It doesn’t cost you anything. You don’t have to kick in anybody’s door. You just have to change a number in the tax code and crime goes down.”
Outside of the huge economic costs imposed by death, worse health, and violence related to alcohol, hangovers alone lead to $220 billion in lost productivity each year in the U.S., or $650 for every person in the country per year. Adding this number to the costs of drunk driving alone means that alcohol externalities are costing society over $1000 per year per person. Each person you know is effectively paying $1000 per year for alcohol to be a normal part of our culture, over and above the actual cost of drinks.
Why you should stop drinking even if alcohol will not harm you personally
Because alcohol is killing the same number of people as COVID each year, I think we should treat it as a similar emergency. A difference is that many (though not all) deaths due to alcohol were caused by someone who chose to begin drinking, while each COVID victim did not choose to get COVID. I think that there is a moral difference between the two, but not so much that we should “live and let live” and accept the number of alcohol deaths. In this section, I’ll give some arguments for why we have an ethical obligation to denormalize alcohol rather than taking a live and let live attitude to other people drinking.
Most of the time, people behave in patterns they learned from the culture around them. Almost all of my behavior and what I consider normal was learned and copied from the example people around me set. It is very hard to step back from your social context and modify your behavior. In some sense, each individual drinker is choosing to drink, but in another sense, they are each doing what we all do most of the time: following the example of the culture around them. There have been some studies on how much of an effect the general social environment of drinking has on people, and it seems to be quite a lot. This study suggests that the social example of other people accounts for a sizable amount of each drinker’s alcohol consumption:
How malleable is alcohol consumption? Specifically, how much is alcohol consumption driven by the current environment versus individual characteristics? To answer this question, we analyze changes in alcohol purchases when consumers move from one state to another in the United States. We find that if a household moves to a state with a higher (lower) average alcohol purchases than the origin state, the household is likely to increase (decrease) its alcohol purchases right after the move. The current environment explains about two-thirds of the differences in alcohol purchases. The adjustment takes place both on the extensive and intensive margins.
We can compare the normalization of alcohol to the normalization of other behaviors. During the pandemic, part of the reason I was wearing a mask and socially distancing was to set an example for other people. If I lived in an area where no one was wearing a mask or distancing, it would have felt more difficult to choose to wear a mask and distance. Even though each person was individually choosing to mask and distance, our collective choices had a strong effect. We understood that we had a responsibility to normalize a specific behavior to keep everyone safe, even though each person was ultimately responsible for their own actions. I argue that we have the same responsibility with alcohol.
We can think of ideas and behaviors as spreadable in the same way diseases are spreadable. Before vaccines, we understood that we each had a responsibility to be extremely careful about not spreading COVID, to the point of going months without meeting people indoors. The idea that “it is normal and good to consume alcohol” is harming more people each year than COVID did, and it takes much less of a sacrifice to avoid spreading the idea than it takes to avoid spreading COVID.
I don’t think it is much worse to be murdered than it is to die of a harmful highly addictive personal habit. You are a victim of outside forces in both cases. The fact that alcoholics are mainly harming themselves rather than being harmed by other people bears little moral weight for me. In both cases, they deserve a culture that protects them. Denormalizing drinking is one way we can protect them.
1 person dies from alcohol each year for every 1900 people who drink. This means that throughout a lifetime of drinking (let’s say from 20 to 70) 1 person dies from alcohol for every 38 people who regularly drink. We can think of the 38 people as each casting a vote to normalize alcohol and fund the alcohol industry. Because it takes so few votes for an additional person to die, I do not want to be part of that 38.
60% of alcohol is sold to the top 10% of drinkers, who drink an average of 74 drinks per week. This number is misleading and skewed by the top 1% of drinkers, who drink much more than the top 10%, but the fact remains that the majority of drinks are sold to the tiny minority who consume them the most. When you step into a liquor store, the majority of the alcohol you see will likely be sold to people whose health it’s significantly harming, some of whom are addicted to it. When you pay for alcohol, you’re paying a business that is being significantly funded by addicts of the substance it’s selling. You’re helping it stay in business and grow.
I have a special responsibility as someone who is not genetically predisposed to alcoholism to avoid normalizing behaviors that would ruin the lives of people who are not as lucky, in the same way, that we have a responsibility to donate resources to people who were not given as many opportunities as we were.
In elite culture (which in many ways I consider myself a member of) there are sometimes disturbing status rituals where you engage in more dangerous behavior to demonstrate that you’re strong and rich and able to bounce back. There’s pressure to do more intense drugs and to drink more heavily. It’s hard not to see this as a way of filtering out people perceived as unfit or weak. While most people I know do not participate in this, I have been in social spaces where these rituals happen. They’re a form of social Darwinism and I don’t want anything to do with them. Concern about alcoholism is sometimes seen as an implicit weakness, and more people not drinking and raising concern about alcoholism can make it socially easier for even more people to avoid alcohol.
Alcohol is so normalized that most people have no reaction to negative statistics about it. It’s culturally understood that alcohol causes problems, so it’s easier to ignore how harmful it is. If alcohol were anything else and had the same negative consequences, it would be clear that we should not participate in it. If Minecraft were killing 140,000 Americans every year, and if 40% of intimate partner violence occurred after the perpetrator was playing Minecraft, I would not buy or play it. Like alcohol, Minecraft is very fun, but it would not be worth it for me even if I knew it would never personally harm me, because it was having so many negative effects on other people. I would not want to financially support it or normalize playing it.
I can’t think of another activity that’s both as dangerous as alcohol and as normalized. Smoking is more dangerous but much more frowned upon. Driving is more normalized but less dangerous.
Because so many people are being harmed by alcohol, even very small changes in drinking habits can lead to lots of lives being saved. This 2015 study suggests that an alcohol tax that raises the price of alcohol by 10% would save between 2000–6000 American lives per year. If a six-pack of Bud Light costing $0.50 more would save the same number of Americans every year as the number who died on 9/11, it seems clear that individuals setting an example for their peers can also have an outsized influence on the number of people harmed.
It will always be unclear how much effect the example we set can have, but I think it’s higher than we assume. We are highly influenced by the lifestyles of the people we’re closest with. If the 5 closest people in your life were heavy drinkers, you would probably have a very different relationship with alcohol than if your 5 closest people did not drink. Change in behavior of one person in a social group can have unexpected effects, especially if that person is well-respected and clear about why they decided to not drink.
Conclusion
In the past, I’ve been greatly benefited by alcohol. Being drunk sometimes left me so happy that I lingered in a more positive emotional state for weeks after. I take the social benefits of alcohol seriously, but I don’t think they’re worth the cost of so many vulnerable people being harmed and killed, and life without alcohol is just as exciting and fun for me. I understand that for many people giving it up would be more of a sacrifice, but it seems clear that the sacrifice is worth it to protect the people alcohol would otherwise kill and harm.
Thanks for cross posting my blog post! Wrote this 2 years ago and would probably update it a bit, one commenter pointed out a mistake:
"1 person dies from alcohol each year for every 1900 people who drink. This means that throughout a lifetime of drinking (let’s say from 20 to 70) 1 person dies from alcohol for every 38 people who regularly drink."
I'd really like to see your source on it because I feel like breaking it down to 1/38 every year in a lifetime of drinking misinterprets the data.
I'd like to circle back and correct a few other points but I stand by the broader claims.
One other point I'd make is that I definitely don't consider harms from alcohol to qualify as a high impact intervention for EAs to think about. It isn't especially tractable or neglected and there are much bigger problems.
Sweeping social taboos are unlikely to work when most people experience net positives, or the negative effects are hard to attribute. See veganism for an example here, and tobacco for a counterexample (many people experienced large, visible net harms).
I can see good arguments for policy work in neglected countries, and I like the work that Concentric Policies are doing here. But a broad social taboo in high income countries seems really, really hard.
Broadly, I think the argument might also be somewhat unnuanced. You could achieve similar effects on alcoholism with targeted policies, which makes it different from veganism and tobacco, where almost all incremental units of consumption cause harm.
(But again, super loose thoughts, open to critique on this)
There’s a paradox where countries with high taxation on alcohol also have high consumption—but this might just be correlative
I expect it to be causation in the other direction. Slamming on the brakes might be correlated with accidents, but it's neither correct to say brake-slams cause accidents nor that the correlation is spurious and has no causal element.
(My comment is only about that specific point and not about anything else you said)
US public is extremely not well versed in ketamine, but I feel low doses of ketamine (20- 60 mg) feel quite similar to 1-3 units of alchohol (same loose feeling, less buzz, less hangover) - although effects diverge greatly soon there after in dose.
I think this is a good thought. With loneliness and social capital underdevelopment such large and apparently consequential problems, it is important to think about alternate candidates that might perform alcohol's social lubricant role
This post does a good job of highlighting the harms from alcohol.
However, I'm strongly suspicious of the implicit framing:
If you talk either about crime policy or drug policy, that’s got to be the number 1 recommendation — just because it’s so easy. It doesn’t cost you anything. You don’t have to kick in anybody’s door. You just have to change a number in the tax code and crime goes down.
This is a quote - rather than the author - but I think the article does the same thing.
Namely, that it takes a very naive view of the subject by focusing on the immediate harms/ tangible harms whilst ignoring the more diffuse and harder to articulate benefits.
The argument could ultimately be correct, but this style of reasoning makes me very nervous, it's like you're setting yourself up for a fall. "It doesn’t cost you anything!" - oh, not in monetary terms it doesn't, not in monetary terms!
That's a good point in general, but I'm less worried about this in the context of an increased tax. The potential magnitude of the loss of benefit is limited by the ability to pay the tax and continue at current levels of consumption ~ in that case, the loss to the drinker is limited to the amount of the tax increase. So I don't think we have to worry as much about calculating the "more diffuse and harder to articulate benefits" as precisely as we would under a ban.
Moreover, you could pair an alcohol tax hike with a decrease in some other consumption tax in a way that makes the whole package ~cost-neutral for moderate drinkers.
On the contrary, I think we should be very careful about imposing morality taxes. I'm not going to say we should never impose them, but not even attempting to think through the unintended consequences is the height of arrogance. I see this bad both from the perspect of leading to bad policy and also bad from the perspective of class relations.
Good point. If rulers worry about the consequences for social life, for instance, they could reduce taxes on some other good that is important to bar and restaurant operations, or even reduce taxes on low alcohol% beverage options while raising taxes on stiffer blends
One can also imagine monetary costs being inflicted on families whose drunk adults now have less money leftover from their binge for picking up takeout or groceries
I think Chris is pointing to the idea that the benefits of alcohol (and other targets of 'this style of reasoning') might be real, even if they aren't monetary. e.g. for alcohol, imagine there is major positive effect where alcohol leads to more socialising. This isn't easy to represent monetarily, but that doesn't mean it isn't real. The style of reasoning he is pointing to is one where you've already measured the downside, and then wave away the upside.
Alcohol taxes seem like the go-to intervention from what I can tell. Kilian et al. 2023 reports that doubling alcohol taxes reduces alcohol consumption by 10%. But I'm unsure the causal-correlational mix of the evidence. When I've looked at less comprehensive reviews of only natural experiments, they come off as more equivocal.
Does anyone know of any organization's focusing on advocacy for alcohol taxes?
Reset Alcohol is an initiative by Vital Strategies, which has received a $10mil grant in 2024 (following $15 million in 2022) from Open Philanthropy. They reportedly contributed to the passage of a recent alcohol tax in Sri Lanka. Concentric Policies appears to be focusing on tobacco control -- presumably because they / AIM thought it was more tractable?
Idk about magnitudes but in my country, Ecuador, alcoholism is a huge problem. We have some sin taxes on alcohol, but it seems difficult to raise them in part because binge-drinking is such a popular passtime and in part because there is lots of illegal production of moonshine-type alcohol that would likely gain market share with higher taxes on formal-sector alcohol products
Does anyone know if satellite and AI technology authorities could more easily identify acreage dedicated to sugar cane cultivation and tax this land more heavily than land devoted to other crops? Sugar cane is an important input for much of the bootleg moonshine in Ecuador, to my knowledge. Sugar cane also produces other harmful products, so I think taxing its production in this way could have other helpful effects and lead to more land and other agricultural resources being devoted to healthier crops
One note: I am sceptical that hangovers should be counted as negative externalities:
Outside of the huge economic costs imposed by death, worse health, and violence related to alcohol, hangovers alone lead to $220 billion in lost productivity each year in the U.S., or $650 for every person in the country per year. Adding this number to the costs of drunk driving alone means that alcohol externalities are costing society over $1000 per year per person. Each person you know is effectively paying $1000 per year for alcohol to be a normal part of our culture, over and above the actual cost of drinks.
Assuming functioning labour markets, much of the effect of hangovers on productivity seems like it should be internalised by the worker. If you show up to hungover, you will be less likely to be promoted, more likely to be fired, receive smaller bonuses etc.
(I agree you are right in cases where labour markets do not function as well and reward is less well tied to productivity, like with trade unions or government employees)
It would be fair to apply a downward adjustment before converting into an externality. At the same time, a worker's salary is only a portion of their productivity, the correlation between productivity and wages may not be particularly strong, and some of the costs nominally borne by the worker end up being borne by society (via lost tax revenue, increased demand on need-based social service programs, etc.)
(As an aside, the article is paywalled, but I'd need more convincing on the $220B figure. I quickly saw a study in the Netherlands that suggested a cost there of 2.56 billion euros [or roughly 60 billion if you scaled to the size of the US economy]. Not suggesting that is the right figure either, but this strikes me as a case in which the methodological assumptions could make a big difference.).
via lost tax revenue, increased demand on need-based social service programs, etc.
I am skeptical of this style of argument, because it seems like it allows a sort of illiberal rhetorical transmutation. A government can take some private aspect of life that it does not have the right to regulate, subsidize part of it, and then claim that those who behave in undesired ways are 'demanding' social assistance, negatively affecting the taxpayer, and hence can be regulated.
I expect the causal effect to be pretty weak - if some light drinkers become nondrinkers, I guess that would lead to some reduction in the amount of heavy drinking, but not very much.
There are larger social influences from "the 5 closest people in your life [being] heavy drinkers" or from social rituals that actively pressure people into heavy drinking, but if the recommendation is for light drinkers to become nondrinkers that doesn't directly touch either of those causal pathways.
People often neglect to compare the costs/risks to the benefits (which I think this post largely does).
In the case of alcohol, I think the benefits of alcohol are smaller than most people believe.
I think in the majority of cases, the feelings of happiness from alcohol are not actually coming from alcohol, but from socializing, then being misattributed to the alcohol.
One easy way to verify this: try drinking alone. You'll find it doesn't make you happy. It just accentuates existing feelings.
Then, try talking to people for awhile without drinking. You'll find once you get past the initial awkwardness, you start feeling good.
Some added nuance: I think alcohol is short term beneficial for getting into social situations, by decreasing fear. However, I think this usually becomes a crutch. It's better to simply learn how to get into conversations without alcohol. It's possible to develop this skill without having to ingest a neurotoxin.
One easy way to verify this: try drinking alone. You'll find it doesn't make you happy. It just accentuates existing feelings.
try doing acid and running a marathon. Try doing molly and going to work. Trying taking Adderall and doing nothing. Drugs don't need to be strictly better in every life situation to be sometimes directly positive. In fact I think your argument is perhaps a better argument for the potential utility of drugs that is not currently achieved. We do a terrible job educating children on how much of different drugs will be fun and what situations it will be most fun. Yet people's revealed preferences are that they still value them highly.
Alcohol can absolutely be fun. It can absolutely be useful for creating social connections and letting loose a little bit even if you could do without. I say this as someone who thinks alcohol is one of the worst drugs for sure by cost/benefit (on a personal level) and broadly agree with the overtures of this post (at least on the negative side, it think its possible alc is extremely beneficial for our culture but hard to say)- I barely drink and would choose most other options.
I think you are taking a very rat/aspie perspective. There are tons of idiots with 0 social anxiety who frickin love shotgunning beers. They are not doing it to feel comfortable with their best friends on a fishing trip. It's fun. at least sometimes.
I saw this on Substack and thought it was great. I stopped drinking almost 6 years ago in part to try to show solidarity with a close family member who is an alcoholic. It sadly did not have the affect on that person I was (naively) hoping for, but I have noticed a uniformly positive reaction from my friends and acquaintances. Most of them want to drink less.
One guy in particular, who I wasn't even close with, approached me at an event recently. He was so excited to tell me about how he had cut his drinking way back. I had never encouraged him to do this. He just knew that I had quit and he clearly felt this newfound connection with me. That was nice.
I agree that there are social benefits to drinking. I have many fond memories of having beers with friends and having incredible conversations full of laughter. My biggest worry in quitting was that I'd miss out on this bonding time. I thankfully have been able to almost fully recreate these experiences, just over a coffee instead of a beer.
Also, the version of me from 15 years ago probably would not have had the courage to talk to my now wife at a party had I not been drunk. But once you do a few things sober you realize life can still be very fun (and even more fun the next morning!) Dancing at a wedding sober and still having an amazing time was a huge milestone, after that I felt like I could do anything.
Thanks for sharing, Kat! Thanks for the post, Andy! I estimated the following (expected) reductions in life expectancy:
For a man (woman) consuming 10 g of alcohol, 16.6 (33.2) person-min.
For the consumption of 1 g of sodium, 0.558 person-min.
For the consumption of 100 mL of sugar-sweetened beverages (SSBs), 5.47 person-min.
For driving a car 100 km in the UK, 6.62 person-min.
For the consumption of 100 g of unprocessed red meat in the United States (US), 9.06 person-min.
For a man (woman) smoking 1 cigarette, 17 (22) person-min.
So I think a small beer with 200 mL containing 5 % of alcool, which would have 10 g of alcool, decreases life expectancy roughly as much as 1 can of Coca-Cola (330 mL of SSBs), driving a car for 300 km, 150 g of beef, or 1 cigarette.
I often tell friends and family about the loss in life expectancy caused by the consumption of alcohol, and unprocessed red meat, and smoking tobacco. My personal experience is that people do not update their consumption much as a result. I feel this is explained by some scepticism about the risks (particularly of unprocessed red meat), discounting of future welfare (which is egoistic), but also by genuinely different preferences. Sharing information about the risks of supposedly harmful consumption empowers people to make the best choices for them. So I think it is generally better to increase human welfare than banning or taxing such consumption.
At the same time, Irecommend funding the Centre for Exploratory Altruism Research’s (CEARCH’s) High Impact Philanthropy Fund (HIPF), which makes grants to decrease the consumption of SSBs and salt, to increase agricultural land, and therefore increase the welfare of soil animals (given my best guess that they have negative lives).
Several commenters noted that moderate drinking can make casual conversation easier to start and maintain. A related concern I have is that completely abstaining from social drinking might create modest but meaningful barriers to building relationships and networks.
I’m not suggesting that the highest-impact career connections happen at bars, but alcohol does seem to play an important role as an accessory in some personal contexts relevant to professional life. While this doesn’t apply to every setting, I worry that what might appear to be a small restriction on a social behavior could have marginally negative impacts on building both personal and professional relationships, which could aggregate over time and/or across many individuals in communities like EA. Given the value of those professional relationships, I worry that general abstinence from drinking could be moderately costly. I think that’s an important consideration given the popularity of this post.
Two things I’ve read recently influenced this take are Peter Wildeford’s Forum post about “weirdness points” and Nick Cooney’s book Change of Heart, both of which I thoroughly enjoyed and found valuable. I think there’s a broader argument to be made for the value of generally conforming to popular social customs without strong justification to deviate.
That said, I haven’t spent the time to put numbers to this concern, which makes me hesitant about its strength and made me hesitant to post this comment. If I’m wrong, I think a likely reason is that I’m overestimating the social relevance of engaging in moderate social drinking or the extent to which it’s common in relevant contexts. I eventually decided to post this comment because I put credence in the heuristic I proposed in the former paragraph and because I didn’t notice this specific concern raised among the other comments (many of which I found very insightful).
Another thought on this post: Usually when I see a title like the one on this post in the EA Forum, I take it to mean the author is making a recommendation based on principles of EA prioritization (ITN framework/cost-effectiveness/etc.). Andy Masley’s comments indicate that despite his view of the legitimacy of the general arguments behind the recommendation in the post’s title (“you should probably stop drinking”), the arguments don’t meet the bar for EA prioritization. I realize the title is taken from the original blog post, but I worry that its context as a post title on the EA Forum may lead some to assume that the poster is claiming it does meet that bar. I understand this kind of criticism could seem like making a mountain out of a mole hill, but I think it points to important posting norms related to the predictability and reliability of information I would expect many readers consume quickly when reading the Forum homepage.
The positive effects of consistent, strong, and healthy relationships are so much more powerful than other factors, they become essentially rounding errors in the equation.
I overall agree that alcohol is a net negative to society. But if it is ingested in moderation as part of social bonding rituals like evening meals together or spiritual services, the positive effects of these rituals on our health—especially bolstering our amazing immune systems that constantly repair mild to moderate damage in our bodies—more than make up for the negatives of the alcohol.
