From the comments, on alcohol abuse

I refer you to Prevalence of 12-Month Alcohol Use, High-Risk Drinking, and DSM-IV Alcohol Use Disorder in the United States, 2001-2002 to 2012-2013. My apologies for not being able to locate the primary data sooner.

Key summary quotes below:

Twelve-month alcohol use significantly increased from 65.4% in 2001-2002 to 72.7% in 2012-2013, a relative percentage increase of 11.2%

The prevalence of 12-month high-risk drinking increased significantly between 2001-2002 and 2012-2013 from 9.7% to 12.6% (change, 29.9%) in the total population.

The prevalence of 12-month DSM-IV AUD increased significantly from 8.5% to 12.7% (change, 49.4%) in the total population.

Twelve-month DSM-IV AUD among 12-month alcohol users significantly increased from 12.9% to 17.5% (change, 35.7%) in the total population.

At the end of the day, I am still going to trust outcomes data over survey data. People lie, autopsies don’t. What I know is that acute alcohol poisoning increased by 700% in 20 years. You die from acute alcohol poisoning not because you slowly got sick over years, but because you drank so much so quickly that your body is overwhelmed. And this is in spite of the medical profession getting better at hemodialysis to bring down acutely toxic ethanol poisoning.

What I also know is that alcohol related hepatic deaths bottomed out in 2003 and have since been rising rapidly (~50% increase). This is due to the fact that the generation socialized by prohibition had lower lifetime alcohol use and problematic alcohol use than the generations before or after. As that generation died off, or aged out, successive generations who drank more started refilling the hepatic wards. Even more fun for every age bracket, we are seeing more alcohol related hepatic death than we saw a decade ago for those same age brackets excepting only the youngest cohorts.

These are basically impossible to square with a thesis of no substantial change in drinking patterns. They fit quite nicely with formal epidemiological surveys showing more problematic drinking and a shift in alcohol consumption.

That is from “Sure,” see also his/her other comments in the longer thread.

Comments

I looked up what 12 month use meant. From Wikipedia...

Alcohol abuse was defined in the DSM-IV as a maladaptive pattern of drinking. For its diagnosis, at least one of the following criteria had to be fulfilled in the last 12 months:

Recurrent use of alcohol resulting in a failure to fulfill major role obligations at work, school, or home
Recurrent alcohol use in situations in which it is physically hazardous
Recurrent alcohol-related legal problems
Continued alcohol use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of alcohol.

So these aren't one offs, not someone binges then doesn't do it again.

One substantial difference in drinking behaviors are the disappearance of drinking establishments here. There are different liquor licenses, from a pub and bars which serves alcohol , a restaurant that serves food and alcohol together. The pubs and bars where you could go for drinks have almost all closed down because of the vigorous prohibition against drunk driving and the prohibition of smoking in public premises. Night clubs and sports bars which offer drinks and some other activity are doing fine.

Talking to the people who run these places they tell me that the drinking clientele now drink at home. Delivery services are available. Instead of drinking in a public place with their peers and the bartenders who would keep an eye on the extremes, they are alone with their bottles.

There is a substantial proportion of the population who will seek some kind of intoxicant. It could be that for the people doing it alone is very dangerous.

'One substantial difference in drinking behaviors are the disappearance of drinking establishments here' and ' because of the vigorous prohibition against drunk driving' are a good example of a culture shift - would it have been normal to go to a bar, drink 4 beers over 3 hours, then go home? Sounds not unusual at all for a number of 20 year olds in 1980, and yet today, that would be against the law in the U.S.

It is fascinating to see how increasing strictures do not seem to lead to the desired results, but that has marked America's relationship to alcohol for a long time.

It’s hasn’t achieved ALL desired results, only some.

What has marked Americans’ relationship to many things, including the afterlife, is an anti-conservative utopianism. “Arguing and hectoring, leading to moral improvement”, is such a feature of American life since ~1600 that it’s our greatest unique achievement, second only to the professional sports-entertainment complex.

'that it’s our greatest unique achievement'

Not even close.

'second only to the professional sports-entertainment complex'

Ever heard of this obscure thing called the World Cup?

Another possible benefit for autonomous cars: bringing drinking back out in public (while also reducing drunk driving deaths).

That only sounds snide: If there is a benefit to drinking, beyond the possible cardioprotective benefits of 1 glass a night, it has always been social bonding. Going out and having a few with your friends is entertaining when you do it and conducive to the sort of social bonds that make for a rich life. But it's pretty hard if you live in the suburbs.

Yes, taxi services exist, but they're expensive and annoying. No one wants to ride an Uber every night.

At least in more populous areas, Uber and Lyft allow people to go ouf drinking and not worry about driving. I'm not a big drinker but if I am going out and want to drink without concern, and I can't bike for some reason, then Uber it is.
Autonomous cars by the way be be horribly expensive; they won't be an option for people who don't have a fair good income.

A previous comment was seemingly eaten by MR's system (so apologies if duplicated):

Full data and definitions here - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5710229/
" high-risk drinking was defined as drinking 4 or more standard drinks (a drink equals 14 g of pure alcohol) on any day for women and as drinking 5 or more standard drinks on any day for men. In this study, high-risk drinking was defined as exceeding the daily drinking limits at least weekly during the prior 12 months.

An individual was considered to have a DSM-IV AUD diagnosis if the respondent met criteria for alcohol dependence or abuse in the past 12 months. A diagnosis of dependence required 3 or more of the 7 DSM-IV dependence criteria, and a diagnosis of abuse required 1 or more of the 4 abuse criteria. Respondents with a 12-month abuse or dependence diagnosis were classified as having 12-month AUD."

Sample size two cohorts of face-to-face interviews, run by "National Institute of Alcohol Abuse and Alcoholism", about 40k people each (80k total), $90 compensation, invite.

Gallup polling which I cited to Sure (https://news.gallup.com/poll/1582/alcohol-drinking.aspx) is about 1k in any year (so about 12k over a comparable 2002- 2014 period), telephone polling.

I'm not "Sure" myself what the superior sampling methodology is here or why there would be such contrasting results.

The above suggests "between the early 1990s and the early 2000s, 12-month alcohol consumption increased from 44.0% to 65.4%" prior to their stated increase of "alcohol use increasing from 65.4% (95% CI, 64.3%-66.6%) to 72.7% (95% CI, 71.4%-73.9%)".

This is in contrast to Gallup showing more or less stability at about 60% since 1940 with a spike in the 1970s. There's an even larger difference here!

I'm not sure how to take the contention that population level frequency of drinking has increased by 100% since the 1990s. Does this seem insane to anyone who was actually an adult in 1980-1990, or is it just me?

Moreover it's the opposite of Sure's original contention in our discussion, which was that population level teetotalling had increased since the 1990s, hence flat total alcohol consumption must be buoyed by increases in use by the hard core users....

Sure said he/she trusts autopsies over the survey data he and you quoted. They found binge drinking death to be growing a lot over 2 decades. Yes, drinking too much is bad. A quick check shows 2200 deaths a year. That certainly is 2200 more than necessary but it is relatively tiny compared to say the complications of type 2 diabetes.

It is probably comorbid with diabetes, with one condition increasing the risk of the other, and vice-versa.

The data on autopsies is presented like there was an increase in deaths by 700%. But from what base? And what percentage of people are autopsied? Like many anti alcohol people Sure is engaging in motivated reasoning. Cars for instance kill and injure many people, but driving is permitted rather than requiring people to only use public transport. Of course alcohol can also do harm, but the vast majority of people do just fine with it.

Numbers are hard to find.

Graphically this figure at Popular Science, using "deaths for drug overdoses" for alcohol sourced from CDC data suggests 1.8/100,000 in 2010 and 0.6/100,000 in 1999 - https://www.popsci.com/science/article/2013-04/which-drugs-actually-kill-americans/

An increase of another 400% of 1999 would get you up to Sure's 700% increase and 4.2/100,000. Overall drug related overdose is up around 225% from 1999-2010, against 300% for alcohol.

Comments often get eaten if there is a link that the filter does not approve of. A duplicate comment will show a message that the text has already been posted-

As ChrisA above and Axa and dearieme note, fashions and trends in diagnosis of cause of death and autopsy, and changes in techniques, may be enough to "fool by randomness" medics looking at rates, particularly if they have a motivated bias to look for it.

Besides this, it is probably a bad idea to generalize up from a rare and unusual form of death, which is probably underdiagnosed a fair bit as an intentional form of suicide, to population level alcohol use in last year rates, or consumption rates.

Do we really trust survey respondents so little that we'd rather try and scale up from a rare and unusual cause of death, a specific behaviour subjects to fashions and epidemiology of suicide, up to a pretty commonplace and socially acceptable form of drug use which most of the population believes is fine?

When we have surveys like the NSDUH with 70,000 sample size per year - https://www.ncbi.nlm.nih.gov/books/NBK424860/figure/ch1.f3/ or the Gallup which show unchanging and reasonably intuitive levels (seemingly to me more intuitive that the suggestion by Sure's NIAAA survey of choice making the strange claim that only 44% of the US population used alcohol in the 1990s, while 73% do today....).

Rare forms of death are quite handy if they correlate well with problem behaviors. Mesothelioma, for instance, allows to track all the other much more common impacts of asbestos exposures. If Mesothelioma rates increase, we can reasonably conclude that the population under observation had elevated exposure to asbestos and related pathology.

We can do the same with coal workers pneumoconiosis and lung cancer.

Because it is not just acute ethanol poisoning that is up. Hepatocellular carcinoma? Up. Alcoholic steatohepatitis? Also up. Acute pancreatitis? Admissions are up. Esophageal cancer? Rapidly increasing incidence last I checked. The incidence of basically every ethanol related disease is increasing or there is some readily apparent reason for it going the other way. For each of the classic alcohol related diseases we see a similar pattern.

Acute ethanol poisoning just happens to have the highest correlation with binge drinking. There is less statistical noise in the relationship. It also shows the effects of small changes in means or consumption distribution can have massively disproportionate changes on outliers under a power law. The single best determinate for how many episodes of acute ethanol poisoning an ED will see is simply how many binge drinkers exist in its catchment area.

And like with so many other things, we can either find special reasons why a cancer strongly related to heavy alcohol use is surging too ... or we can except what every medical student learns - asking "Do you have occasion to use alcoholic beverages such as liquor, wine, or beer?" is a terrible instrument. We have much better designed questionaires in practice and I still routinely get people who will only endorse alcohol use when I personally ask them. From personal experience, face to face interviewing ends up matching the objective tests (blood ethanol levels, transaminase ratios, etc.) much better than simple questionaires.

As far as suicides by alcohol ... no.

Suicide by alcohol is actually exceedingly difficult; I see far too many people who failed in the attempt for it to be that effective. Very few people can guzzle enough alcohol before they before they become too wasted to absorb a toxic dose. It is far more common to see people cause lasting brain damage than for them to actually die from acute alcohol toxicity. A quick view of the literature suggests that completed suicide is so rare that it shows me only case reports.

Everyone I have pronounced dead from alcohol ingestion had multiple bouts of massive ingestion (e.g. >20 standard drinks in a sitting).

Well, you're pretty dug in on your confidence level in inference of inferring alcohol use from diagnosis and treatment rates. I can't say as that's changed my mind about the problems of that and its lack of simplicity and the general problems from working back to patterns from rare events, but such is arguing on the internet (you persuade the gallery of viewers, never ever the opponent, however trivial the proposition).

While I'm here, can I ask your opinion on your initial statement on : "Likewise, the number of teetotalers in the US has increased. Back at peak ~70% of Americans reported that they drank in the last year. Today that number is ~60% (lifetime abstention has gone down, but both "last year" and "last month" have gone up). In 1990 the US consumed 6.5 liters of ethanol per capita, today we consume just over 7 (close to the 70s peak) and are rising."

Would you still support that statement in light of your survey's claim that 12 month drinkers go from 44% in the 1990s, to 62% in the 2000s to 73% in the early '10s?

Ultimately, to get back to the premises of our discussion, we've from my PoV been discussing how small a population may have changed its heavy drinking patterns here, my initial statement being that simple alcohol use hadn't changed and consumption hadn't changed, so heavily increased consumption of sort that Cowen pointed to must be concentrated in a tiny population. Your rejoinder was seemingly that this population must be quite large due to substantial increase in teetotalling (as above and contradicted by the latest source you cite) combined with flat consumption.

If we're back to the point where we are talking about change in a very small population (people who could generate a 0.005% change in overall population causes of death) then thats pretty compatible with my premise in our initial discussion and there's not much more of a discussion to be had from my PoV.

Actually, you will note that I said 12 month use is up, but fewer people are reporting habitual use. And this is the problem with recall data, for people who drink at weddings and wakes they may report themselves as last 12 month or as teetotalers. A similar problem comes in with sacral wine. Some folks will report that as quarterly (mostly 12 month) or weekly use. What the synthesis of all the data looks like to me is that we have seen a rise in minimal use (I drink on special occasions) and in problem use (I drink to get drunk).

What I do know is that the youngest cohorts have been drinking less than they peers did in previous generations. With increased volumes of ethanol being sold and taxed in the last few decades this means somebody else has to drink more. I am seeing more people come in with alcohol related illness. The death stats for alcohol related illness are going up; the less confounded a diagnosis is, the more it is going up. The affordability of an addicting substance has gotten cheaper; a large number of the most problematic drinkers report that they are only constrained by the price of alcohol.

You have a valid data point that Gallup surveys this often and has both different and steady numbers. Unfortunately it is not replicated (Pew, for instance, shows different use rates with similar methodology) and more importantly many other sources of data tell a different story.

Looking at the totality of data suggests that problematic drinking rates are up. The health sequelae are up. Barring some new change, the trendlines are cause for concern.

How much acute alcohol poisoning involves kids who don't know their limits bolting down shots until they drop? I never got that bad in my misspent youth, but I can recall some nasty vomit sessions and a couple of epic hangovers before I realized when enough was enough. If these deaths are mainly young'uns going off the deep end then I doubt they correlate very well with trends in longer term alcohol abuse.

'Suicide by alcohol is actually exceedingly difficult'

So, single driver car fatality with high blood alcohol level - how would you classify it?

Or is simply not possible to classify reliably, so it remains unclassified even as suicide?

Or how about a single person firearm death involving alcohol - generally easier to classify as a suiclde, but what role did the alcohol play?

That would not be acute ethanol poisoning. If you can turn on the car you basically do not have acute ethanol poisoning.

People with acute alcohol poisoning cannot protect their airways. Very fundamental reflexes that we share with all land mammals stop functioning. People who kill themselves by MVC, GSW, or whatever are virtually always have insufficient alcohol on board to get an acute alcohol poisoning ICD-10 code.

If we want to go with alcohol being listed in Part I or Part II sections of the death certificate, sure my gut experience is that suicide secondary to alcohol intoxication is up. But when you go for the ICD-10s with the cleanest relationship to binge drinking, we are up.

The demographic changes in alcohol use in Sure's cited study are worth note (however we take it):

Men: >6.8%, Women: >15.8%. White: >8.3%, Black: >24.2%, Asian/Pacific: >29.1%, Hispanic: >17.2%. 18-29: >9.6%, Older than 65: >22.4%.

I would surprised if there was any change in use by young white men at all (unless a strong version of Simpson's paradox operates). Might even be falling. It seems like any increase in drinking (if not "problem drinking") should be accounted for by ethnic minorities, females and older folk.

It seems like a very rapid period for such intense change in the above demographics. Most of these groups up because at lower baselines (same is true for poorer and less educated individuals; converging towards a higher mean of alcohol use by richer and more educated individuals).

Mirrored by anything noticeable in the culture?

Whites were already at the top for alcohol use and remain at the top according to table 1:

White 69.5 (68.2-70.8) 75.3 (73.7-76.9) 8.3

75.3! Compared to Asians or Blacks that started at lower lower numbers.

Black 53.2 (51.6-54.9) 66.1 (63.8-68.3) 24.2
Asian 48.4 (44.3-52.5) 62.5 (59.4-65.5) 29.1

The Northeast and Midwest are the biggest drinkers. The South drinks relatively less.

Regions don't exist.

Eh, you're looking at the "drinking at all in the last year" numbers at Table 1. I think it's more relevant to look at the "problem drinking" numbers in Table 2, which show Blacks and Hispanics going from below Whites to above Whites, and American Indians always being above Whites but going either higher.

Asian rates are presumably at least partially kept low by East Asians and ALDH2 deficiency, though people with ALDH2 deficiency who drink heavily are more at risk for negative consequences (including cancer of the esophagus.)

I still think my point about obesity rates going up is an underrated factor here.

It's a bit difficult to go to a bar by foot in the suburbs. In a urban setting you can walk or take a short bus/train ride to go home after the bar.

Another case of revealed preferences? Living in a suburb displays contempt for bars =)

Measles is up 300% in one year!! Unlike alcohol, epidemics spread much faster. People forget that things like the flu once sent millions of human beings to early graves.

https://www.washingtonpost.com/world/2019/04/16/world-health-organization-says-measles-cases-up-percent/

'Measles is up 300% in one year'

From a very low number, it should be noted. And it is extremely far from becoming an epidemic (though obviously, the unvaccinated will die at higher rates than the vaccinated).

This is another change from my youth, where vaccinations in Fairfax County public elementary schools completely normal, for example (the vaccine gun was much better - faster and less painful - than needles, in my memory). Or you could simply go to the county public health office in Fairfax Ciity, and also get free vaccinations.

And yes, in that sense, people forget how it used to be a normal function of government to spend money on public health measures based on decades of experience. Strangely, that perspective started changing right around the start of the 1980s, and continuing until today.

The deaths from acute alcohol poisoning cited also starts from a very low number. While one should be mindful of quantities consumed, this prohibition talk lately on at least on this blog has been more a moral panic than a real health scare. For example, opioid continue to kill more than alcohol poisoning but where's the scolding about the evils of painkillers?

That's an easy question:

Encoded into law.

Fifteen states now limit first time opioid prescriptions to seven days. A few more limit it to 14. Multiple states have morphine equivalents per day limits. I believe every state now has formal (prohibitive) use guidelines. Oh and most states are now tracking opioid prescriptions at point of purchase and requiring prescribers to check recent fills prior to cutting a script.

Drunk driving deaths are currently about half the opioid overdose deaths. So you think we should be about 1/2 as stringent on alcohol as opioids?

And here is some other information involving younger people - 'The National Institute of Drug Abuse just released its 2018 report, Monitoring the Future: National Survey Results on Drug Use, 1975-2018, that combines data gathered from about 45,000 eighth-, 10th- and 12th-graders in public and private schools across the United States on their drug and alcohol consumption. The good news is that alcohol use by older adolescents is in decline, and in some age groups, it’s at the lowest rates ever reported.' https://www.washingtonpost.com/lifestyle/2019/06/24/when-teens-drink-its-different-than-adults-heres-how-you-can-help-your-child-avoid-problem/

It is easier to create a narrative (which we all do) then fit facts to match it, then to acknowledge that the reality of what is happening is simply too complex to fit any narrative.

Age plays a role too, because this was striking in another sense - 'That said, the number of teens who engage in underage drinking remains concerning. Nearly 60 percent of 12th-graders and one-quarter of eighth-graders report having drunk alcohol.' Basically all of the 12th graders between 1975-1981 or so could legally drink at 18, creating an interesting shift in the problem of 'underage' drinking.

Definitions seem to have also changed in another sense - 'When teens do drink, they drink differently than adults. Teens are more likely than adults to binge drink, defined as four or more drinks on one occasion for women and five or more drinks over two hours for men.' Since (roughly) one drink an hour is metabolized, if a woman was to drink four glasses of wine during 3 hours of 'one occasion' involving things like playing volleyball and grilling on a sunny summer afternoon, would that be binge drinking ? Apparently, though seemingly not for a man drinking six cans of beer over that same 3 hours.

I make no claims concerning changes in drinking behavior over the years, but sometimes, it feels as if there is simply no straightforward baseline that covers 4 or 5 decades of massively shifting attitudes. What is now decried as a problem sounds pretty much like normal behavior for many 20 year olds in 1980, all of who legally purchased the beer or wine they were drinking.

Which is why the baseline problem is real - it is quite possible that the sort of drinking not unusual in 1980 has been squeezed in both directions - some people drinking less (after all, in 2019, a 20 year old drinking, much less buying, alcohol is committing a crime), and another group drinking considerably more (helped along by things like noticeable alcohol industry support of events like spring break).

Nonetheless, watching moral panics roll over the landscape is just part of getting older. And at times, those moral panics prevent public health measures from being employed to reduce the harm (the saddest example of this involving the U.S.'s response to AIDS, but America's obsession with alcohol has a much longer track record).

And thinking along the lines of public health, and the use of hepatic, one assumes that Americans are now being offered, as a public health measure, the appropriate vaccines against hepatitis, right?

It isn't as if what is routine in a health care rationing country like Germany is not even more common in the unrationed health care system of the U.S., right?

The prevalences of hepatitis B and C in Germany are quite high for the developed world and exceed those of the US.

You keep trying to create a health-insurance issue, but the numbers aren't there. When I pointed this out in the other thread, you changed the subject and dumped some unrelated paragraphs, as is typical for you. I predict that the same will happen here.

No, you keep thinking this is about health insurance. The point of vaccinations is about public health.

And links are always useful, so people can judge for themselves whether the subject was changed, or some unrelated paragraphs were dumped.

And you might just remember a certain major event involving Germany in 2015, resulting in this - 'From September 2015 to March 2016, hepatitis A notifications in Germany increased by 45% to 699 cases compared to 482 cases in the same period of the previous year.'

And truly, links supporting whatever it is you assert would be nice, particularly as people can judge for themselves what is stated. And again, note this has nothing to do with health insurance, but is about public health - 'In Germany, all hepatitis A cases are notifiable to local health departments irrespective of nationality or status of residence of the case. Local health departments are responsible for case investigations and prevention measures. Individual case data are compiled in a case report, anonymized and electronically transmitted to the state health department and further to the Robert Koch Institute (RKI).' https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5457677/

Here is some data for the U.S. - https://www.cdc.gov/hepatitis/statistics/2016surveillance/index.htm I'm sure you can provide your links for anyone to make their own data comparison, or would that be the sort of thing you consider to be changing the subject?

Kids, if you want to be as smart as Tyler or Alex, don't drink, don't do drugs. This public service announcement is brought to you from the owners of Dixie Cups.

I was reading Nick Maggiulli, a talented financial advisor who was class valedictorian, Stanford grad, straight A student, so he was already smart, who became briefly a binge drinker while in school (it did not affect his grades) due to peer pressure. Brave man to talk about it so candidly, see here: https://ofdollarsanddata.com/just-one-more/

That was good. I also had that experience in HS and college. I don't drink at all. For many years.

I would give up booze and women if I could read books as fast as Tyler. Imagine zipping through the political economy of Burkina Faso in 15 minutes flat. In one month I'd own the whole world!

Plus in just another 15 minutes he's posted a Burkina Faso bleg and toured the place in person

alcohol related hepatic deaths....yes, alcohol is dangerous.

Now do diabetes and all the disabilities caused by it.

This is a fair point, but remember that diabetes impairs kidney and liver function, so many alcohol-related maladies are comorbid with diabetes. Alcoholism exacerbates diabetes, at least until the patient starts insulin therapy, and by then it's likely already too late (for type 2 diabetics).

Could the increase in alcohol-overdose deaths be due to a synergistic interaction with opioid abuse? I.e., opioids + alcohol => liver damage => greater susceptibility to alcohol poisoning?

Also, is a death from taking alcohol and opioids simultaneously always going to be recorded as an opioid overdose, or combined drug toxicity, or do doctors make a judgment call about which is the dominant contributor?

Opioids don't really do anything to your liver (unless you share needles and catch Hep C), so it's unlikely that increased opioid use has much to do with increased deaths from alcohol.

Alcohol and opioids are both CNS depressants, so using them together increases your risk of respiratory failure. Most of what we consider to be opiate ODs are in fact the result of multiple drugs, often including alcohol.

RX opioids (e.g, Vicodin) are very often compounded with acetaminophen which is toxic to the liver in large enough doses. People with serious Rx pain med abuse do suffer liver failure.

Google a death certificate. We can, and do, code multiple direct causes with multiple sequences. And of course there is Part II.

In any event, the big synergy for killing comes from decreased respiration. If you have both on board, opioids are going to be listed if only one is for some reason. The pupils and other tells are going to lead the jury down the opioid track far faster than down the alcohol one.

If anything wonky is going on with death certificates with alcohol and opioid combos, it is going to understate alcohol and overstate opioids.

Everything in our culture seems to be reflected in extremes, from political and religious views to performance in school to alcohol consumption. I've commented before how my generation was the bell curve generation: almost everyone fit near the middle, whether it be performance in school or behavior outside of school. What happened to cause the bell curve to flatten out and have such fat tails? Everyone is so different now.

One explanation is the high level of inequality today. At this blog, that's not an accepted explanation of the cause of fat tails but rather the consequence of other factors. For example, the economy is producing more losers than winners because of significant IQ differences. One may recall the discussion after the financial crisis that the risk curve had fat tails as the result of the high level of inequality, but that discussion has long since passed. Anyway, I don't expect inequality to subside any time soon, so we can expect the fat tail discussion to resurface from time to time.

" high level of inequality today" I don't know. It seems it's the richer white, 'highly educated' that are the wackjobs today. They aren't the ones suffering from inequality - only narcissism.

But how many of them are wrecking their lives and health with booze and drugs? Outside the entertainment industry I don't think you'll find that many. Like it or not the upper and upper middle classes generally do have their acts together in multiple ways. Many of our social problems emerge from the bottom third more or less of the income and wealth distribution. I'm willing to consider inequality (perhaps of opportunity and therefore hope) as a major factor in that, but it's silly to blame well off people for dysfunctional behavior they do not often engage in.

I can see the great advantages of data from autopsies rather than surveys. On the other hand, even the results of autopsies will be influenced to some extent by fashion, won't they?

"People lie, autopsies don’t" seems to me to be a bit too strong.

Only problem is autopsy is not that frequent, could there be sampling problems?

The hospital autopsy rate is below 10% in the USA. https://jcp.bmj.com/content/68/8/601

Good point but it isn't the only problem. Reported Cause of Death is almost certainly influenced by local habits - so that, for example, international comparisons are difficult - and also mere fashion.

I have no idea how often it's influenced by incentives but many human activities are so I wouldn't rule that possibility out.

Did the autopsy rate change? Did the autopsy selection bias(es) change?

Autopsies are imperfect, but those imperfections were baked in back in 2003. It would be an awfully odd change that occurred first in the 90s (driving down apparent alcohol deaths) and then reverting back up in the 2000s.

We are looking for signs of change, so any alternate explanation has to be something that was different in 2003 than either 1999 or 2009.

It's been interesting to see the change in social norms, at least in my slice of life, mostly driven by safety culture. The firm where I worked adopted a zero tolerance for drinking and driving in the 90's; people were fired for what had been common practice a few years before.

It had been, for example, fairly routine to take the team out for beer occasionally. That just stopped, because everyone was driving home afterward. Company parties didn't completely ban alcohol, but it was limited and carefully managed. The hosts for business dinners were similarly responsible for ensuring moderation.

I wonder how the social demand to legalize and normalize drugs will play out vs. the social demand for absolute safety.

But is the experience in your firm common out there? In tech, I see the opposite: Free alcohol in the office is not uncommon in the tech sector: Right next to the flavored sparkling water. It's frowned upon to drink in the morning, but you'll see people start drinking at 4 or so, and earlier on Fridays.

Similar things happen in tech conferences: Nobody goes to the conference to attend the sessions, but to network over binge-levels of drinking in either bars or sponsored afterparties.

This of course doesn't meant that this is common outside of tech either: To see cross-society pictures, we need wide studies, not pictures from small samples of companies.

Nice one Sure, +1.

I've been watching a few Korean rom-coms(*). It's amazing how much getting seriously drunk is part of every story arc(**).

US teen exploitation comedies might have that, but for the rom-com demographic, movies are more likely to go with "drinking is bad."

An interesting cultural difference.

* - for example ‘Abyss’ On Netflix. You might say it's not a rom-com with so many murders, but a lot of Korean rom-coms have murders! And suspicious guys in blue pants, blue shirt, and blue baseball cap.

** - In fact I learned that there is such thing as hangover soup.

In fact I learned that there is such thing as hangover soup.

Korean hangover soup? Talk about a hair of the dog!

It seems to me that girl friends and wives tend to push men to drink less. Could the rise be due to the rise in singleness?

+1

I'm not at all convinced it's true, and in fact wondered if the effect ran in the other direction, but it's nonetheless an interesting question.

Opioids, obesity, marijuana, relationship patterns. They all strike me as important potential confounds. I'm not sure how much obesity changed from 2001-2013, though

It seems to me that girl friends and wives tend to push men to drink less.

Yes, but her brother keeps pouring me shots!

Another very good source of data for alcohol consumption is taxes on alcohol. Pretty much all OECD countries have some sort of specific taxes on pure alcohol. There has been a dramatic shift from wine to liquors.

'There has been a dramatic shift from wine to liquors.'

Or not, quoting more extensively from a source of information that disputes many of the more glib explanations being happily bandied around here - 'Since the 1970s changing alcohol consumption has led to a three- to fivefold increase in liver deaths in the UK, and a three- to fivefold decrease in France and Italy. These changes in opposite directions challenge popular notions of static ‘drinking cultures’ including the idea that in Southern Europe there has always been a culture of harmless light drinking, integrated into everyday life. The radical positive change in Southern European drinking patterns has been explained as part of a wider societal transformation since the 1960s, with growing urbanisation and related changes in working conditions as well as increased health awareness in the population.

In contrast, the change in the UK and other Northern European countries, moving from a lower base in the 1980s towards higher levels of drinking and alcohol-related harm is generally explained by increased availability and affordability of alcohol, combined with a culture of heavy, episodic drinking connected with weekends and celebrations, which are considered to have a particularly bad effect on health.

Therefore, it is important to note that the effect of societal changes on alcohol consumption differ between countries and there is no direct link between changes in living conditions and drinking. The process is indirect, mediated by each country’s historical and cultural background.'

Please do the math to determine which generation was "socialized under prohibition " and tell me how this explains the rise in binge drinking over the past two decades. Rising singlehood and demographic change seem more likely explanations.

Perhaps the increasing severity of drunk driving laws has played a role. Like someone above said, people are more likely to drink greater quantities of booze at home than in bars. Used to be, more people would go out drinking with a plan to drive home. Now they plan to stay home to drink because knocking back a few and driving home is too financially dangerous.

'financially' dangerous? How about just plain dangerous? There's a reason drunk driving is so severely punished.

Sure, but it was always physically dangerous. What has changed is that the law has made it increasingly financially dangerous to get a DUI. Or do marginal costs not matter?

People are driving drunk less but binge drinking more. This gives us a clue as to *where* people are doing more of their heavy drinking now vs. 20 years ago.

Sounds like a law that actually worked as intended. The second order effects (more drinking at home) are not good but I suppose Uber is helping with that these days.

See my post on the rise of imperialism (imperial rather than triple) in the hazy IPA market. You nailed it. High gravity craft breweries serving hardcore high carb haze accessible by uber in hipster enclaves of major cities likely accounts for the increases. Bravo!

Prohibition is a particularly weird one to talk about because alcohol consumption probably recovers by at least 1950 - http://gdpsychtech.blogspot.com/2013/01/estimated-alcohol-consumption-during.html

If not largely by 1945 - https://history.stackexchange.com/questions/34745/did-alcohol-consumption-increase-in-the-u-s-after-1933

So what's the generation socialised under prohibition or even its after effects, here?

We need to compare this data against the rise of imperialism in the niche hazy IPA market. And by Imperialism, I mean the tendency to avoid the triple IPA label. Hazy IPA introduces a huge increase in carbs and compounding basal complexity in Type 2 diabetics. I wouldn't be surprised to see a correlation, much like autism, to proximity to major highway arteries in hipster enclaves of large cities.

"People lie, autopsies don’t."

I, for one, welcome our new autopsy-performing robot overlords.

Apologies if this has already been pointed out, but after 9/11 liquor ads were allowed on television once again, ostensibly to help boost an advertising economy during a recession. If the increases noted began in 2003, that sounds about right.

Comments for this post are closed