ICE is now taking more people than ever. Nearly 50,000 were arrested in July.

Just yesterday, Mullin announced the arrests of 2,197 immigrants in the past month in New York State. The DHS press release trumpeted that these included “murderers, rapists, pedophiles, drug traffickers, and violent assailants.”

But as the New York Times reported, in the past month ICE has instead focused on grabbing immigrants who never had a brush with the law.

In July, according to ICE’s statistics, 73% of people leaving ICE custody had never been convicted of crime, including 45% who never even had criminal charges. Those proportions will likely increase with this new wave of arrests of innocent people.

When ICE grabs people, it locks them up in detention centers. There are more than 200 of these facilities around the U.S., with ICE managing some and contracting to local sheriffs for others. Nineteen detention centers hold more than 1,000 people each. This one in Pine Prairie, Louisiana holds about 800.

A few months back, I wrote about the skyrocketing numbers of people held in these jails. According to the statistics on ICE’s website, it jumped to some 75,000 this past January before dropping back to about 60,000 in the past few months. That’s still nearly three times the number held in 2022.

Last year, ICE tried to expand its detention capacity beyond 100,000 by buying and converting abandoned warehouses, shelling out $1 billion for 11 of them. This year, faced with fierce community opposition and the near-impossible task of converting warehouses into jails, the agency reversed course. According to the New York Times, it’s now trying to resell seven of those 11.

That leaves the active jails increasingly crowded, for example, looking like this:

Spending even one day in a packed, cinder-block tank like this must be a miserable, stressful experience, but ICE detainees are locked up much longer than that. With a backlog in immigration courts, the time that immigrants are held in these facilities keeps growing, from an average of 40-50 days in 2023 to about 80 days in June 2026.

Plenty of people stay longer than the average. Here’s the distribution of the length of stay in June 2026. More than 8,000 people were held in these facilities for six months or more.

It’s difficult to get good information on what conditions are like inside the jails, but besides the crowding, detainees have complained about cold temperatures, insufficient toilets, verbal abuse, and – most consistently - being denied medical care.

Crowded, indoor settings like this create many health risks, from infections to mental illnesses, which I summarized in a post a few months ago. And when you incarcerate 60,000 people, many are bound to have major medical problems. Unless you have top-quality medical providers and sympathetic security staff, those medical problems can become dangerous or even fatal.

And sure enough, the number of people dying while in ICE detention keeps rising. This year, the agency listed 21 deaths just through June. That puts 2026 on track to have four times as many deaths as 2024 and about 25% more deaths than 2025, which was the worst year by far.

That’s not just because the number of detainees is growing. Here are death rates per 10,000 detainees, with my projection of the death rate for 2026 based on that first-6-month toll.

When the rates triple over a few years, it suggests that two-thirds of those deaths might be preventable.

ICE doesn’t say publicly what these people died from. But based on past reviews, we can make an educated guess about the most common reasons. They are, in roughly descending order: chronic diseases (like heart disease and cancer), suicide, violence, and infectious diseases like COVID-19 and influenza. And I think it’s fair guess that these deaths are just the tip of a growing iceberg of disease and injury caused by ICE detention.

Overall, this situation is very ugly. I don’t think it’s right for people in our country to be put through this. But I especially don’t think it’s right if their only crime was to come to the U.S. to escape violence elsewhere, work and feed their families.

Share

Subscribe now

I’ve been raising alarms about marijuana for years. There are many reasons to be very alarmed: use has been steadily rising, as has addiction to the drug. Pregnant women are increasingly getting high, risking damage to their babies’ brains (with what may be a true cause of autism). And drug-induced psychosis is on the rise (here and here). One big reason that I’ve been harping on this is that few others are.

So I follow surveys of marijuana use closely. In this newsletter I try to explain trends in data, but I’ll confess up front that I don’t have a solid explanation for the trends that I’m seeing now.

In the annual National Survey on Drug Use and Health (NSDUH), for about the last 30 years, marijuana use has been falling in teenagers and rising in young adults. The drop in use by teens has been an important talking point for groups favoring legalization. I don’t know why teens are increasingly avoiding a drug that adults are embracing, but I’m certainly not complaining about it.

I expected that teens would eventually use the drug more as they copy their older siblings, friends, and parents. But instead, since 2022, the surveys show marijuana use by adults 18 to 25 years old dropping.

The drop isn’t big - only about 10% - but it’s a very surprising reversal of the trend over previous quarter century.

A similar reversal happened a year or two later in adults 26-29, 30-34, and 35-39 years old.

Adults older than that – age 40 and up – are continuing to warm up to the drug. The graph below is cluttered because I used the same scale as the one above, but you can get the gist: use by folks who are middle-aged and higher is still steadily rising.

The graphs track use of marijuana in any form in the 30 days before the survey. Most of those users are getting high only occasionally, but about one in four uses the drug every day or nearly every day. These near-daily and daily users have the greatest risk of depression and psychosis. Among all young adults, the percent who are near-daily or daily users followed a similar pattern as that for all users: a steady increase from the 1990s to 2022, then a modest drop.

What’s the story behind these strange trends?

The biggest factor driving the long-term rise in adult use is clear: legalization. Probably the best evidence is that the early-legalization states showed the earliest and biggest increases in use.

But then why the reversal in the 2020s? I can think of three possibilities:

Survey artefact. The methods for NSDUH have changed. For many years, the survey relied on in-home interviews, but in 2021 an online option was added. Overall response rates are about 45%. Maybe the reversal has to do with changes in the type of people willing to sit for the survey.

But a completely different survey using consistent methods (the Monitoring the Future project) showed a similar pattern, with a leveling (though not a decrease) of use in young adults starting in 2021.

In this Monitoring the Future survey, there wasn’t a drop in the percent of people who were daily users, though. But this was a much smaller survey, making the year-to-year changes noisy.

Also, annual surveys in Canada (which legalized the drug in 2018) have found a small decrease in marijuana use in 20-24 year olds since 2020.

Post-pandemic recovery. Maybe people used more marijuana during the COVID-19 lockdown period in 2020-2022 and are just recovering from that spike in use. That’s possible, but it seems unlikely. There wasn’t a spike in use of alcohol during the lockdown period, so why should there have been a spike in use of marijuana?

Wising up. This country doesn’t have any policies or programs in place designed to persuade young adults to stay away from marijuana. We should. But maybe by 2022 young adults had experienced or witnessed enough of the ugly effects of regular weed use to decide on their own that it’s best to avoid it. I couldn’t find data to track this in the U.S., but in Canada, the percent of cannabis users who say that the drug had a harmful effect on their mental health rose from 6% in 2019 to 10% in 2024.

With one in 12 young adults in the U.S. getting high nearly every day, almost every young adult must know someone with a weed habit. Many probably know someone who has become addicted, depressed, psychotic, or all three.

If this is what’s causing some folks to take a pass on the drug, you’d expect that the drop in use to be bigger in the early-legalization states, where use has been the highest. State-level results from NSDUH haven’t been posted yet, but regional-level data have, and they do show bigger drops in use in the West and Northeast, the regions with the early-legalization states.^[i]

So maybe the most encouraging explanation is true: young adults are beginning to back off marijuana use because they’ve seen what regular use does to your brain and they don’t like it. That’s only a guess…but a hopeful one.

I welcome any other explanations or thoughts from you readers.

Marijuana use is still an enormous problem in this country. The drug sold now has a potency of THC of 18-23%, which is about seven times the potency from 1980. One in four young adults uses it, and about one in seven meets the criteria for cannabis use disorder. That’s more than 5 million young adults – and 19 million Americans of all ages – that are addicted to the drug. And even as cannabis may be losing its luster among young adults, the drug is increasingly popular among middle-aged and older adults.

But it’s a tiny relief that we may be seeing early signs of a self-correction.

Share

Subscribe now

^[i] California, Alaska, Oregon, Washington, Maine, Nevada, Colorado, Maryland, and Vermont

The CDC is reeling after withering attacks by elected officials, brutal staff firings and resignations, and an extremist shooting. By my count, eight leadership positions are now vacant, and (with one three-week exception) the agency hasn’t had a permanent director for 19 months. Meanwhile, it’s facing rolling waves of measles outbreaks, a food contamination epidemic that has sickened thousands, and the fastest-growing Ebola outbreak in history.

Can Schwartz right the ship?

Who is Dr. Erica Schwarz?

Dr. Schwartz opened her testimony at the Senate confirmation hearing this way: “The most important thing to know about me is that I come from a God-fearing, military family.”

And sure enough, the military is a strong theme in her life story. Her father was in the Navy, and three siblings are in the Navy, Marines, and Army. After finishing an 8-year combined BS/MD program with the help of a Navy scholarship, Schwartz spent seven years in the Navy and thirteen years in the Coast Guard. Her specialty is occupational medicine, that is, keeping sailors and coast guardsmen safe and healthy.

In the hearing, Dr. Schwartz came across as the consummate military officer: intelligent, respectful, well-spoken, and poised. But, for someone jumping into the maelstrom of the Trump administration, she was also shockingly oblivious.

Most people I know are glued to news feeds these days, bracing for Trump’s or Kennedy’s next outrage. But when Schwartz was asked about the elimination of CDC’s Office on Smoking and Health, she said she was “not aware” of the problem.

In the public health world, the elimination of that office, and the parallel firing of the director of the FDA’s Center for Tobacco Products was (and still is) a major news story. It’s hard to fathom how she missed it, especially with her calling smoking cessation a major focus of her work in the Navy.

By now, we’ve become accustomed to Trump appointees lying to Congress. But here I believe Schwartz was telling the truth. That makes me wonder what else about the trashing of the CDC she’s missed.

Granted, Schwartz was trying to pick her way through a minefield. The senators wanted to pin her down on whether she would refuse Kennedy’s or Trump’s orders if she believed they would hurt the public’s health. Her answer, over and over, was that she didn’t believe those bosses would issue harmful orders.

Of course, she was obligated to say that. If she didn’t, Trump would pull her nomination.

Maybe I’m too gullible, but her response struck me as also being honest. It looks like Schwartz truly believes that Trump and Kennedy wouldn’t force her to do things that would damage public health.

And that’s very, very strange. The previous CDC Director Susan Monarez was fired precisely because she resisted Kennedy’s attempt to take vaccines away from children. If Schwartz believes (as she said repeatedly), that as CDC Director she would “never betray the science”, shouldn’t she have asked in her job interview whether Kennedy would overrule her on vaccines or other important health decisions? Is it conceivable that she didn’t ask that? Does she think it’s inappropriate to question an order from her boss?

Her most telling response was to a pointed question from New Hampshire Senator Maggie Hasan. The Senator asked whether Schwartz agreed that the CDC Director was, according to the charter of the Advisory Committee on Immunization Practices (ACIP), at “the top of the chain of command” for vaccine recommendations. Schwartz answered that the CDC Director could accept or reject the ACIP recommendations, but then added, “I understand that – and I can be wrong on this – that the Secretary ultimately has the policy decision regarding that.”

She’s a good soldier. She follows the chain of command.

What would she do at CDC?

Today, CDC needs - for starters - an internal leader who will support the embattled staff, hire the best available experts, earn trust, and develop a clear plan to fight measles, Ebola, foodborne outbreaks, and those chronic diseases that Kennedy claims to care so much about. And an external leader who can translate science-based facts for average Americans.

But there is more to the job than that. CDC is at its best when no one knows what the hell is going on. That is, the agency rises to the challenge when a new pathogen appears (think legionnaires’ disease, AIDS), or an old pathogen takes a strange turn (think monkeypox), or a self-inflicted crisis crops up (think prescription opioids). Responding to those kinds of crises require original thinkers - people who can look at raw data, ask difficult questions, reason for themselves, and challenge conventional wisdom.

From what we know now, is Schwartz up to that?

On the plus side, she’ll likely bring order, structure, and standard processes. She’ll try to fill empty staff positions. She’ll support the troops.

On the minus side, she’ll probably look for Standard Operating Procedures when they don’t exist. She may struggle responding to unknown unknowns.

Most important, I’m afraid that when Kennedy tells her to do something wrong, she’ll follow orders.

Erica Schwartz might have been a fine CDC Director in an earlier era, when HHS secretaries were too preoccupied with our medical care system to pay attention to CDC. But when the HHS Secretary is hell-bent on forcing his unhinged views about health on the nation, a good soldier can be very destructive.

What will this mean for public health?

In Washington today, public health policy is set by crazies and cronies. The most obvious example is the recent Trumpian vaccine policy, issued by an Executive Order that contradicts law, science, and common sense.

To protect us, health experts are refuting or just ignoring what the federal government says. Look at how quickly Trump’s vaccine order was rejected by infectious disease experts, the American Academy of Pediatrics, and state health departments.

So the loss of Americans’ trust in federal public health leadership is now complete. I don’t know what it will take to regain that trust. But this new CDC Director, in this administration, can’t do it.

Share

Subscribe now

Now suppose you’re a community leader in Iowa, and you’re trying to battle the twin epidemics of obesity and type 2 diabetes. Would you: A) post videos on social media about what food to eat, or B) push for limits on the marketing of junk food?

Because you’re reading this, you might choose option B in both scenarios. That would put you in a small minority, though, because in the obesity and diabetes scenario this country overwhelmingly chooses A. And the result of that choice is that we are sicker than we ought to be.

The media has decided: health is up to you

The media that we all consume every day is overloaded with health advice. For example, Healthline.com weighs in on protein choices and “Biblical Eating”.

Even the august New York Times offers tips on psyllium husk and vaginal estrogens.

And social media gushes with advice. A recent study by the Pew Research Center turned up 6,800 “health and wellness” influencers with more than 100,000 followers each. More than 500 had over a million followers. Forty percent of U.S. adults (and 50% those under 50) say they get health information from social media influencers.

The tips are mostly about fitness, weight loss, personal appearance, and mental health.

Some of that advice is misinformation – lies to sell something useless (like magnets or beet juice) or clickbait to sell advertising.

But most of the advice that I see is more-or-less scientifically accurate.

Still, the totality of this advice is a danger of its own. The problem is not in the scientific accuracy, it’s in the framing. The starting point for health tips is that it’s all on you. You are in control of your health.

That framing resonates with us as Americans. Our founding story is that, with enough moxie and hard work, individuals can overcome any obstacle. Therefore, if you just follow a few tips every day, you can avoid flab, or dementia, or diabetes, or cancer. It’s a prevention strategy for our times, the age of the selfie.

Even good advice misses the mark

To what degree does our health actually depend on things that we as individuals control?

The most important health advice today is incredibly simple: avoid smoking, using alcohol or taking other recreational drugs; eat less junk food and more fruits and vegetables; and get regular exercise. According to estimates by the Global Burden of Disease project, between 1% and 12% of Americans’ years of life lost can be attributed to not following elements of this advice.

Taken as a group, these five behaviors are responsible for about 25% of Americans’ years of life lost.^[i]

Another way to come at this is to ask what percent of the disparities in mortality rates by income or education can be attributed to these behaviors. The answer is the same – about 25%. Medical care might explain another 5% of these disparities.

That means at least 70% of our health is outside of our individual control. I say “at least” because even these key behaviors are much harder to control than most people think.

That 70% isn’t just genes or luck, though. Most of it has to do with the environments that surround us, like the quality of the air we breathe, the chemicals in our water supply, the products on store shelves, the price of groceries, the dangers on our streets, or the images that flood the media. As individuals we can’t change those, but as a society we can.

The most obvious example of bad framing: obesity

One evening about 20 years ago, while I was browsing in a large bookstore, I counted the number of diet books on the shelves. The total: 142. At about that time, about 70% of adults said that they wanted to lose weight or avoid gaining weight. Since then, obesity rates have risen by about a third. Today 72% of American adults are obese or overweight. Does anyone really believe that more diet advice will solve this problem?

The obesity epidemic is caused not by individual ignorance or irresponsibility, but instead by the marketing of junk food. By marketing I mean using all of the “P” tools: designing hard-to-resist products, pricing and packaging them attractively, placing them within arm’s length, and promoting them heavily with feel-good advertisements.

Sure, each of us can control over what we eat in a single day. But it takes willpower to resist this marketing of junk food, and most of us are too preoccupied with the other hassles in our daily lives to maintain that willpower every day for a lifetime.

Nonetheless, the media continues to frame obesity as the result of individual people making bad dietary decisions. A study of press stories between 2005 and 2016 found that 46% framed obesity as an individual problem, 23% as stemming from the environment, and 31% as caused by both the individual and the environment. Between those years, the percent of stories using the individual frame increased.

Today, Americans are primed to believe that infectious diseases are outside of their individual control. The explosive diarrhea of cyclospora comes from contaminated lettuce, which Americans blame on the lettuce growers or the government. COVID-19 travelled through the air, so Americans (at least initially) supported social distancing rules to slow the spread. Americans mostly see people with HIV infection as victims of forces beyond their control. But the overwhelming view of a non-infectious, chronic condition like obesity is that it’s the person’s fault.

In fact, people have more individual control over cholera than they do over obesity or type 2 diabetes. During a cholera epidemic, handwashing can cut your risk of cholera by about 70%. But having a relatively healthy diet reduces your risk of obesity by about 50% and developing type 2 diabetes by only about 30%. We know that relying on handwashing is the wrong strategy to stop cholera epidemics. When will we learn that giving out diet tips is the wrong strategy to combat obesity and diabetes?

Will the addiction label help?

Plenty of experts have tried to persuade the media to pin the obesity epidemic on the food industry’s marketing. One route to get there is to label junk food as “addictive” – ergo, beyond a person’s control. The common understanding that nicotine is addictive helped pass anti-smoking rules.

It’s worth a try, but I’m not sure that label will overcome the dominant individual frame for food. Some experts have been calling junk food addictive for years. The textbook “Food and Addiction” came out in 2012. The journalist Michael Moss published a popular book on this five years ago.

But the tsunami of diet tips keeps coming.

And the other diseases killing us…

The problem of individual framing of problems that are beyond our individual control is much bigger than obesity. Most of the diseases that kill or sicken us today – like heart disease, stroke, chronic liver and kidney disease, mental illness, and most cancers – are preventable, not through individual health tips, but through changes to our environment. Good luck finding information on the internet about that, though. Instead, you find the equivalent of tips on exactly how to wash your hands to avoid cholera. Hot or cold water? Bar or liquid soap? Air-dry or towel-dry?

And that’s why I plug away with this newsletter. It’s one place to discuss not what you can do for your health, but what we can do for our health.

Subscribe now

Share

^[i] It’s inappropriate to just add attributable risks to get an overall estimate, because that would double-count deaths that could be prevented by avoiding any of multiple risk factors.

Here’s an update on a modern epidemic that hasn’t gotten enough attention: depression in young people.

Between 2011 and 2021-22 there was a near tripling of the percent of teens and young adults who were so depressed that they couldn’t function for at least two weeks. The tip of that terrible iceberg was a ~30% rise in suicide rates in young people over the same years. I’m convinced, as are many others, that this epidemic was caused by the combination of smartphones and social media – hardware and software in young people’s hands 24/7 that made them miserable.

Last year, though, I was excited to discover that the trend had begun to reverse itself, with depression falling since 2021 in teens and since 2022 in young adults. Now, the source of these stats – the National Survey on Drug Use and Health – has been updated with 2025 data. The downward trend has continued, showing a roughly 30% drop in depression since those peak years.

That continued decrease was paralleled by a drop in deaths from suicide in older teens and adults through age 34.

Granted, these markers of mental illness are still way too high, but the turnaround in the past few years is a very welcome surprise.

I’m sure that some of the spike in depression can be explained by the COVID disruptions in 2020-22, which threw us all for a loop. But the epidemic started long before COVID arrived and the fall has continued long after the lockdowns ended, which says that other, longer-term factors are responsible.

What are they? When I wrote about this last year, I noted that young people were using smartphones as much as (or more than) ever, but that the way they were using smartphones had changed. Today teens and young adults are spending less time comparing themselves to others on Facebook (which makes their own lives seem dreary) and more time watching videos on TikTok (which may not be educational but isn’t depressing).

Maybe that’s all that is behind the continued trend, too. Or - a more hopeful explanation - maybe teens and young adults as a group are learning how to avoid the mentally-toxic effect of smartphones, spending less time boasting, snarking, and doomscrolling and more time getting together in real life.

Middle-aged adults seem to be the trailing group in both the rise and fall of this trend. (Check out the age 26-49 line in the top graph above.) I now wonder: do those young people have something to teach the rest of us about how to avoid being driven crazy by those devices?

Share

Subscribe now

Which would you choose if instead your option was one hundred thousand dollars tomorrow or two hundred thousand dollars a year from now?

And now which of those would you choose if you were struggling to make the rent payment each month?

The differences in how people answer these questions may help solve a riddle that has bugged me for a very long time: why are poor people more likely to smoke?

The relationship between smoking and income is startling. People living in poverty are five times as likely as those making $200,000 a year to light up.

It’s not just poverty, though. The relationship runs across the entire range of income. Middle class people making $75,000 - $100,000 per year are twice as likely to smoke as those at the top of the income range. With smoking - by some accounting - the leading killer in America, this pattern is incredibly important to our nation’s health.

There’s a similar pattern for junk food. People with less income have much junkier diets.

It’s not at all obvious why people with less money have these unhealthier habits. Here are possible explanations, some of which we can quickly discard:

• Awareness of health risk

Could lower-income people just be less aware of the health risks of smoking?

Nope. In a phone survey of more than 2,000 people, knowledge of specific health risks of smoking was no different in low, medium, and high-income people.

Lower-income people are a little less aware of the health risks of ultra-processed food, but not by enough to explain their eating habits.

• Cost

Are low-income people just voting with their wallets, choosing lower-cost options?

No, especially for smoking. Going through even a half pack of cigarettes a day costs $1,825 per year. The alternative – not smoking – is free.

And as I discussed in a previous post, contrary to what I often hear, healthy food is not substantially more expensive, and may be less expensive, than unhealthy food.

• Access

Could people in lower-income neighborhoods just have greater access to tempting tobacco and junk food?

They do. In most but not all studies there are more stores selling cigarettes in low-income neighborhoods. And low-income neighborhoods tend to have fewer supermarkets and more corner stores offering mostly junk food.

I see these as problems and I’m a big believer in making neighborhood environments healthier. But it doesn’t seem likely that differences in access could fully explain the huge differences in people’s habits. For example, low-income people usually travel outside of their neighborhoods to shop for food, which means that nearby stores may not be very important to their food choices. And a major national study of food purchases found that less than 10% of the differences in people’s food choices were explained by the neighborhoods in which they lived. In that study, even when shopping at the same stores, people with less income bought less-healthy food than people with higher incomes.

• Coping with stress

Some argue that lower-income people smoke to cope with stress. There is evidence to back this up: on surveys roughly 40% of smokers say they smoke to relieve stress, and levels of subjective stress increase as incomes go down. Some might argue that eating junk food is just a way of coping with stress, too.

But I’m a little skeptical of this argument. In most studies, exercise reduces stress, but people with less income aren’t more likely to exercise. It may be that the stress-coping argument is just getting close to the real reason, which has been thought up mostly by economists: hyperbolic discounting.

• Hyperbolic discounting

All of us are often forced to choose between getting something today and getting something better tomorrow. Should we buy that pair of pants now or wait for it to go on sale in a month? Trade in our old car for a newer one or put the money in a 401k? How people make those decisions depends on how much they “discount” the future compared to the present, so it’s called “time discounting” or “temporal discounting”.

Most of us aren’t rational about this: the farther away a time period is from us, the less we tend to care about it. For example, when people are asked if they would rather get $50 today or $100 a year from now, many choose $50 today. But if the choice is $50 in five years or $100 in six years, almost everyone chooses $100 in 6 years – even though the one-year delay before getting the higher payment is the same. Valuing future time less than near time is called hyperbolic discounting. The greater the hyperbolic discounting, the more people live for today, not tomorrow.

Studies from around the world have consistently found that people living in poverty are more likely than people who have money to discount the future. For example, in a survey of more than 40,000 people in the U.K., when given a hypothetical choice between receiving £45 in three days or £70 in three months, people living in poverty were much more likely to choose the almost-immediate but lower payment.

People living in poverty are much more likely to take out payday loans, in which they borrow a few hundred dollars to spend immediately but then must repay much more just a few weeks later.

Some might think that a “live for today” attitude is what made people poor, but many studies suggest that poverty itself may cause people to be biased toward immediate rewards. This has been called the psychology of scarcity. Researchers, for example, have shown that people are more likely choose immediate (versus delayed) rewards just before a payday, when their financial situation is at its worst, than afterwards.

This may seem irrational - some have called it the “cognitive burden of poverty” - but a bias toward today over tomorrow can be very reasonable. If you’re about to be evicted for overdue rent, it makes sense to take out a usurious payday loan (or grab your $100,000 raffle winnings, in the poll above) to pay that rent, even if doing otherwise sounds better financially over the long term. That long term doesn’t look so great if your family becomes homeless.

In fact, there is an entire mathematical theory (called bankroll management) about rational financial decision-making when you have a limited amount of money. Not surprisingly, the less money you have in your bank account, the less money you should risk. And risking less in a financial sense typically means spending rather than investing money.

Unfortunately, even if the habit of choosing short-term benefits over long-term rewards makes sense financially, it’s harmful when it spills over into behaviors that matter to health. People who discount the future are more likely choose the immediate pleasures of nicotine, other drugs, and junk food over the long-term health benefits of passing them up. The result is that they’re stuck with chronic diseases, from obesity and diabetes to lung cancer.

What does this mean for how we get healthier as a nation? Answering that question would take more words than I want to use in this post. But, because the relationship between income and unhealthy behavior seems to be a fundamental one, it does argue that we would be healthier if there were less of a gap between the rich and the poor.

And it also argues that those of us in the health business should listen a little more to economists, who have figured out a thing or two about how we humans behave.

Share

Subscribe now

And that made me wonder how big that toll is today, and how big it might become.

Climate change…

During the 2020 presidential election, nearly every Democratic candidate described climate change as an “existential threat”. In the Trump era, though, the topic has all but disappeared from our national conversation, pushed aside by the threat to our democracy and by political consultants telling Democrats to shut up about it.

But as environmental activist Bill McKibbin points out, “physics is running the show, and it won’t be long denied.” In 2024, global temperatures were 1.2 degrees Celsius above the 20^th century average, and the rise is accelerating.

That average temperature increase comes with more frequent heat waves and other extreme weather events like droughts and sudden torrential downpours. And of course, melting polar ice caps leading to rising sea levels.

…and health

The World Health Organization has called climate change “the single biggest health threat facing humanity”. Those are strong words.

So far, though, the WHO doesn’t have the numbers to back them up. In 2014, a group of experts convened by WHO estimated that between 2030 and 2050, climate change would lead to about 240,000 additional human lives lost annually. When I read their report, I thought: that’s all?

I don’t mean to be callous. A quarter million people dying unnecessarily is an enormous human tragedy. But it’s a relatively small number in a world of 8 billion people, in which 7 million adults die from tobacco use and 2.6 million mothers and children die from malnutrition each year.

There’s no question that climate change will be devastating to other species. It may cause as many as one in 12 to go extinct. But humans are the planet’s most adaptable species, who inhabit settings as harsh as Antarctica and the Sahara desert. Based on what we are seeing now, WHO’s comment sounds like the kind that overstatement that undermines the credibility of people arguing for climate action.

But as I looked into this, I realized that predicting what global warming will do in 20 or 30 years is a fool’s game. Here’s what we know now:

• Heat

With global warming, we are having more heat waves, and heat waves are lasting longer:

Extreme heat increases the number of people who die from heart, lung, and kidney disease. Heat waves are also bad for pregnant women, increasing the risk of preterm birth and stillbirths, and they stoke mental health problems, from depression to aggressive behavior. Best estimates are that heat waves cause 1,000 – 2,500 extra deaths a year in the U.S., of which maybe a third can be attributed to human-induced climate warming.

Worldwide, the WHO experts estimated that between 2030 and 2050 the number of deaths from excess heat would rise from 40,000 to 90,000 a year.

• Drought

Heat and changing weather patterns are increasingly causing droughts across the world, affecting about 10% of global land area.

Drought means that farmers produce less food. That likely won’t affect people like you and me, but it will cause malnutrition in people who survive on the crops they grow, particularly in sub-Saharan Africa. The WHO experts estimated that falling crop yields would cause 95,000 additional deaths a year from malnutrition.

• Floods

A warming planet means that more water evaporates from the ocean, the atmosphere holds more water, and rainstorms become more intense, all of which creates flash floods. Floods can kill people suddenly (and terrifyingly) – as they did in Texas’ Camp Mystic last year - or the floodwaters can have lingering indirect effects by damaging infrastructure and upsetting ecosystems. One review found a 26% increase in mortality in the few months after floods.

I couldn’t find estimates of the number of total deaths caused by flooding in the U.S. or worldwide, but because floods tend to affect limited areas, the total number of deaths is probably not large. (That is no comfort if one of those is your child, though.)

• Air pollution

I’ve long thought that air pollution doesn’t get enough attention as a health risk, maybe because pollution levels are much lower in the U.S. today than in the past. The fine particles and gases in air pollution, when breathed in every day for years, cause damage not just to the lungs but also to many other organ systems. Today, air pollution is thought to be responsible for 80,000 deaths a year in the U.S. and nearly 8 million worldwide, making it a leading killer globally.

The main sources of fine-particle air pollution in the past in the U.S. have been coal-fired power plants and vehicles that burn gasoline and diesel fuel. But in the past 20 years, the combination of drought and heat has set forests on fire, and the smoke from these fires (that dull haze that many of us saw last week) is becoming an increasingly large contributor to air pollution.

With global warming, wildfires are likely to get much more common. Here are recent models of wildfire smoke emissions from the western U.S. under different climate change scenarios. Looking at these maps makes me wonder if we’ll ever see a clear summer day or there will be any trees left in the West by 2060.

The brunt of the smoke will be felt by people in western states. But as we saw last week, wildfire smoke drifts across entire continents, so the largest number of smoke-related deaths will happen in the densely-populated states of the Midwest and Northeast.

In these models, some 41,000 people in the U.S. die per year from the effects of wildfire smoke today, rising to 75,000 by 2050. That’s orders of magnitude higher than the deaths from the heat itself.

• Infectious diseases

It’s easy to imagine how, by disrupting ecosystems, climate change might expose humans to many different infectious diseases, both old and newly discovered. Changing temperature and humidity will allow some micro-organisms and the species that they depend on to thrive and find new hosts to infect.

The biggest single worry today is malaria, the parasite for which needs warm temperatures to spread. Malaria already sickens 280 million people and kills about 600,000 a year worldwide. When cooler areas warm up, the WHO guesses that malaria deaths will increase by another 60,000 per year. It’s possible that malaria could start spreading in the U.S.

With increased temperatures and water shortages from drought, people have less clean water for drinking and washing, which allows diarrheal diseases to spread. That’s not likely to be much of a problem in the U.S., but the WHO estimates that it might cause an additional 48,000 deaths per year in the poorest countries of the world.

Overall, though, infectious diseases are the joker in the deck. When a new virus like SARS-CoV-2 with only 10 genes can cause 15 million deaths worldwide, it’s clear that small changes in infectious agents or environments caused by climate change can have massive human health effects.

The future

So as I look at it, in the next 10 years or so, the human toll from climate change is likely to be substantial, but not catastrophic.

But after that, who knows? We’re just beginning to see the health effects of today’s temperatures, and under different scenarios, global temperatures could rise another one to two degrees Celsius. That will radically shock the ecology of the planet. It is impossible to predict the combined and domino effects of multiple ecosystem changes.

The example of wildfire smoke makes the point: heat + drought => wildfires => air pollution => heart and lung disease. Today that chain of events is obvious, and it’s the effect that is killing the most humans by far. In 2014, though, the WHO experts didn’t even mention it.

The diseases that ultimately kill the most humans from global warming are probably ones that we aren’t even thinking about today. And I’m not placing any bets on how many will die.

Share

Subscribe now

And then we took some swings at soda. Not surprisingly, the big soda companies swung back. Reviewing how this fight played out can offer lessons for today’s battle over ultra-processed foods.

Push and pushback

Before 2000, people tended to blame obesity on individuals’ overeating or high-fat foods. But around that year, it became clear that obesity was a population-wide epidemic and that sugary drinks - while not the entire cause - were major contributors.

Studies increasingly showed that sugar-sweetened beverages were not (as dietitians often described them) “empty calories” but instead were actively toxic. Swallowing a bottle of liquid sugar overloads liver cells with fructose, causing fats to accumulate in the liver and blood, and floods the blood with glucose, which prompts the body to release insulin and store those fats.

Armed with this information, health advocates from school nurses right on up to Michele Obama took on sugary drinks in several ways:

• Protecting kids in school

In 2000, many school districts across the U.S. had lucrative multi-year contracts that obligated them to sell soda to kids through vending machines in cafeterias and hallways. Believing that schools shouldn’t make kids sick, people in public health pushed state and local laws to banish those machines.

The soda companies (and the schools benefiting from the sales) pushed back, with furious lobbying in city councils and statehouses, combined with proposals for voluntary guidelines. The truce that finally emerged is a disappointing middle ground. Under the federal law passed as part of Michele Obama’s Let’s Move, schools can’t sell sugary drinks during regular school hours. But they can still sell diet soda in high schools all day and sell full-sugar soda in all schools after regular hours - which means that soda vending machines still sit in many schools, theoretically locked up from 8 to 3.

• Educating people

Most people get most of their health information, directly or indirectly, from the mainstream media. One way to nudge people to change risky habits is to just generate news stories about the problem. In the 2000s and 2010s, health advocates worked hard to make noise about the connection of sugary drinks to obesity, with some success. Stories in newspapers that mentioned “obesity” and either “soda” or “sugar-sweetened beverages” surged between 2000 and 2016.

After that, though, the topic dropped from the headlines, eclipsed by the opioid crisis and COVID-19. The one exception is the blip in 2025 as states got waivers to exclude sugary drinks from the SNAP program.

The soda companies didn’t leave the messages unanswered. They continued to advertise their full-sugar drinks like crazy, constantly reminding people that, for example, that drinking Coke makes you happy. There is plenty of reason to worry that these pro-soda ads are the only messages that people have been seeing for the past 10 years.

Meanwhile, the companies also invented or acquired new drinks that didn’t carry the weight-gain stigma of regular soda. Some had artificial sweeteners instead of sugar, but others just included faddish ingredients to sugar water so that the companies could sell them as healthier – ingredients like electrolytes, caffeine, vitamins, botanicals, probiotics, protein, and even fiber.

• Counter-ads

Much of the success in fighting smoking came from paid television counter-advertising – those spots that connect smoking with its ugly consequences, from holes in the throat to lost teeth. These ads hit harder than the news stories do, and if they are novel or outrageous, provoke their own press attention.

A few cities have tried counter-ads against soda, and they’ve worked. But these campaigns are expensive, and health departments never had the steady funding needed to sustain them, I suspect because appropriators in Congress and statehouses didn’t want to pick a fight with Coke and Pepsi.

• Taxes

If the 2000s was the decade of awareness-raising, the 2010s was the decade of the soda tax. Or so it seemed mid-decade. Between 2015 and 2018, many U.S. cities considered and seven passed taxes on sugar-sweetened beverages. These taxes worked exactly as health economists predicted. By raising prices, they led people to buy less or switch to low- or no-calorie beverages. More exciting, cities passing taxes have seen a relative slowing of the increase in body weight in kids.

But the soda giants mounted ferocious resistance, with lawsuits, lobbying, and funding of front groups and political campaigns. They mostly won. While sugary drink taxes are spreading in other countries, no U.S. city or state has enacted one since 2018.

• Warning labels

A great way to tell people that a product is bad for them is to put that message right on the product’s package. Graphic warnings on cigarette packs have helped cut smoking rates in other countries.

Singapore now requires that beverages carry “Nutri-Grade” labels like this one, with grades C and D used for those high in sugar.

Compared to the nasty cigarette pack warning, those Nutri-Grade labels are very tame. But after they appeared, sales of C and D beverages in Singapore plummeted by 44%.

In 2020, the City of San Francisco passed a law requiring warning labels on sugary drinks, but the soda companies blocked them with a lawsuit. The FDA could require these kinds of labels nationwide, but the soda companies have had too much sway with the federal government for that to happen.

• SNAP restrictions

In 2011, New York City tried to prohibit the use of SNAP benefits to buy sugary drinks, and the USDA turned it down. After that, no other city or state tried this idea until RFK Jr. resurrected it last year. Today 23 states have USDA-approved waivers to do it.

The soda companies are going after these waivers with an army of lawyers. Recently a judge ruled in favor of the companies in a lawsuit involving five states. The waivers in other states look vulnerable to similar lawsuits.

The score

Before people worried about the health effects, sugary drink consumption was skyrocketing. Between 1965 and 2002, Americans’ soda intake roughly doubled, rising to deliver 21% of our total calorie intake. (The WHO recommends that sugar from food and beverages should be below 5% of daily calories.)

I couldn’t find any surveys of Americans that tracked sugary drink consumption consistently over the decades, but the surveys that are available hint that 2000 was the high water mark. Here’s sugary drink consumption in New York City, in a survey that ran annually from 2007 to 2020:

And here’s a survey of high school students nationwide, conducted in odd years from 2007 to 2023.

In both surveys, it looks like consumption fell from 2007 into the 20-teens but leveled off around the time that the press attention fell.

Since 2014, soda companies have been reporting national data on beverage sales, which show a slow fall. Over ten years, with people shifting to bottled water and diet drinks, they are buying 10% more beverages but taking in about 15% fewer calories per day.

So in the past 25 years, while we’ve made some progress, Americans are still taking in an average of about 170 calories per day of sugar in liquid form; many are consuming far more than that. That’s way more than is healthy.

Considering how awful the obesity epidemic is (half a million deaths per year), and how unnecessary sugary drinks are, it’s shameful that we haven’t made more progress over two decades. This should have been an easy win. But the story shows how big, powerful businesses can thwart even popular efforts to protect our health.

Unless something dramatically changes how this country makes decisions on health, this is how the war over ultra-processed foods will likely unfold. Battles over school food, media messages, labelling, and taxes, involving not just health facts but also advertising, lobbying, lawsuits, community mobilization, and political campaigns. And progress that is achingly slow.

Any movement that hopes to protect Americans from ultra-processed foods needs not just the facts in its favor, but also the weapons and the stomach for this kind of fight.

Share

Subscribe now

Ok, I’m biased by having spent years as a government technocrat. In those years, I often bumped into a fundamental disagreement with elected officials. They believed that, unlike “unaccountable bureaucrats” (e.g., me), they represented the true voice of the people because they answered to voters. I believed that as a salaried civil servant I was unbiased, but they were conflicted (if not corrupted) because they actually answered to campaign donors.

In this battle, last week’s Supreme Court decision in Trump vs. Slaughter gave a decisive victory to the electeds. It’s a dreadful decision, but – oddly - it also offers a sliver of hope for those of us who believe that Americans’ health shouldn’t depend on the paid-for whims of a political party.

The Slaughter

The greatest zealot for the elected officials’ point of view is OMB Director Russell Vought. Early in this term, Vought boasted that he was “embarked on deconstructing the administrative state” and replacing it with loyal party members.

He’s succeeding. Kash Patel is purging the FBI of career staff and investigating Trump’s political enemies. EPA’s Lee Zeldin has gutted the agency while pursuing “energy dominance”. Secretary of Transportation Sean Duffy is threatening to withhold Congressionally-approved funding to local governments for roads and bridges if they don’t fall in line with Trump’s immigrant deportations.

The small pieces of the federal government that should have been protected from party politics are the independent regulatory agencies. Congress established agencies like the Federal Trade Commission (FTC) and the Securities and Exchange Commission (SEC) to regulate complicated parts of our economy in a non-partisan way. These agencies answer to commissioners who by law represent both parties and who are appointed for years that overlap presidential terms.

But Vought went after them, too. Last week the six Republican judges of the Supreme Court gave him a huge win. Under Trump vs. Slaughter the President can now fire those commissioners and replace them with his political loyalists.

So we can now expect men as loyal and slimy as Kash Patel in charge of investigating plane crashes (NTSB), preventing stock market scams (SEC), and inspecting nuclear power plants (NRC). Yikes.

The one exception that the justices made to the Slaughter decision was the Federal Reserve System, which runs the nation’s central bank. The Fed has a complicated leadership structure that keeps it out of the President’s control, which has served the country well for decades. The Fed has prevented bank runs and runaway inflation, and blocked presidents from forcing down interest rates just to goose the short-term economy.

When Trump tried to fire Fed board member Lisa Cook, he was met with a howl of protest from nearly 600 economists, as well as headlines like this one:

At least, the Court refused to let Trump fire Cook. I guess the justices felt that the economy was just too important to leave up to politicians.

I agree. But health is important, too. The difference between life and death is a very big one.

The CDC

While Trump loyalists have been deconstructing other federal agencies, Robert Kennedy, Jr. has been doing the same to the CDC. He’s fired expert members of advisory panels, revised the vaccine recommendations to meet his whims, canceled work that offends campaign donors, and installed political loyalists to run the agency.

So we need a Fed for public health – that is, an expert agency to look out for the health of Americans that is insulated from the whims of the president.

The Supreme Court’s Slaughter decision rested on the argument that, under the Constitution, any federal agency that has “executive power” must answer to the president. Although the justices refused to define exactly what “executive power” is, they said that the FTC has it because it issues and enforces regulations.

With that, the justices left an opening. “If Congress wishes to establish independent agencies to assist it with its functions, it may do so [as long as the agencies do not have “executive power”].”

In fact, CDC has always operated with little to no “executive power”. The legal authority for public health in this country lies mostly with state health departments. CDC doesn’t mandate vaccinations, require lockdowns, or order contaminated items off shelves. It doesn’t issue or enforce regulations. CDC gathers data, analyzes it, and makes recommendations that are not legally binding.

Because it is run by experts rather than politicians, people have listened to CDC. Early in the COVID-19 pandemic, Americans trusted medical advice from CDC by more than 2:1.

That trust remained high through the entire pandemic, despite the disruptions caused by the response. But trust in CDC tanked in 2025 when Kennedy started trashing the agency:

Boards of Health

We already have a model for a public health organization that is insulated from politicians: the Board of Health.

Boards of Health were set up in England in the 1790s to help respond to epidemics of typhus. Government officials battling a killer disease felt they should get advice from the most knowledgeable health experts. The idea spread across the Atlantic, and by the early 1800s many major cities in the U.S. had created their own local Boards of Health.

Local Boards of Health, led by health experts, still operate throughout the U.S. today. Some make legally-binding rules and some are advisory only.

The Trumpian slaughter of the CDC means that we need a National Board of Health, established by Congress and structured like the Fed to be insulated from politics. For example, some members could be appointed by the president and confirmed by the Senate, and others appointed by the Association of State and Territorial Health Officers, with all serving terms that overlap presidential administrations.

The Board could do nearly all of what CDC does now: collect ongoing data, investigate disease outbreaks, and issue recommendations and guidelines on how to prevent diseases. Those recommendations could cover topics a far-ranging as vaccines, food, recreational drugs, and preventive services delivered by health care providers. The Board wouldn’t have “executive power”, but as we’ve seen, it mostly doesn’t need it to protect our health.

In a funny way, this would somewhat follow what Russell Vought himself wanted. Project 2025, which Vought was heavily involved in, proposed that CDC be split in two, one piece of which would collect, analyze, and publish epidemiologic data, and the other of which would make public health policies. A National Board of Health would serve in the first job, only adding nonbinding recommendations to its work.

And it’s not a new idea. In 1879, Congress established a National Board of Health to deal with the leading killers of the day, such as cholera, typhoid, and yellow fever. The Board had 11 members, including seven appointed by president and confirmed by Senate, three medical officers from Army, Navy, and Marine Hospital Service, and one representative from the U.S. Attorney General.

Unfortunately, this board died four years later, being bureaucratically out-maneuvered by the Marine Hospital Service, which handled quarantine. In 1912, that Marine Hospital Service became the U.S. Public Health Service, and in 1995 the U.S. Public Health Service was bureaucratically swallowed up by the Department of Health and Human Services.

Which shows that government bureaucracies change. They are born, reorganized, subsumed by other agencies, and killed off. The Trump administration has damaged the CDC almost beyond recognition. When sanity returns, the structure for carrying out its work needs to be rethought. With the Fed as a model, resurrecting a National Board of Health could protect us from health threats, and from Trumpian politicians.

Share

Subscribe now

I know you already have too many public health risks to worry about, from Ebola to autism, but there’s another one that deserves your attention. This health risk doesn’t involve micro-organisms or toxic chemicals, so you may see it as more of a social problem. But, as with all social problems, this one wind up as a health problem.

The gambling tsunami

When I was a kid, you had to go to the racetrack or fly to Las Vegas to gamble legally. Then came off-track betting and state lotteries. Then casinos on Indian reservations, on gambling boats, and in cities hungry for tax dollars. Then video poker warrens. Then in 2018 the Supreme Court struck down the federal ban on betting on sports, which created “sportsbooks”.

Then a tsunami happened when gambling moved online. Today, Americans can place bets – mostly on sports – 24/7 through computers in their pockets. Sportsbooks are now legal in 37 states, almost all of which also allow online betting.

With this constant availability, the “handle” (dollars bet) on sportsbooks surged 10-fold since 2019.

The sportsbooks now keep more than 10% of the dollars bet (up from 7% in 2019), so $165 billion in bets gives them revenue that tops $16 billion a year.

Gambling companies spur on this online betting with tricks that get users to stay on the sites longer, bet more often, and place bigger bets. These tricks take advantage of the deep information that the apps have about users’ habits, preferences, and vulnerabilities. For example, companies pitch in-game bets, like whether a favorite player will score next. They offer “parlays” (multiple linked bets), nudges like “You have a free bet expiring today!”, introductory bonuses for new users and “reload bonuses” to users whose activity drops.

To draw in more users, gambling companies run huge advertising campaigns, including ads like this one with Kevin Hart and LeBron James.

Beyond sportsbooks are prediction markets (mainly Polymarket and Kalshi), which also involve betting on sports, rebranded as investments. Prediction markets have grown more than 100-fold just since 2024 and now handle billions in bets every month.

The losers

Researchers measure problems caused by gambling with surveys that ask gamblers nine questions, such as “Have you bet more that you could really afford to lose?” and “Have you gone back on another day to try to win back the money you lost?” People answer on a scale of 0 to 3 points for “never”, “sometimes”, “most of the time”, or “always”. Those who score 1-4 points are considered low risk, 5-7 points moderate risk, and 8 or more points “problem” gamblers. (Our President would call them “suckers and losers”.) Most problem gamblers meet psychiatrists’ definition for gambling addiction.

Most surveys that are available on gambling were fielded before the online tsunami, so they underestimate the problem. But surveys conducted in New Jersey in 2015 and 2021 show the early effects of online wagering. Between those years - but before online gambling had fully taken hold - the proportion of adults who admitted to risky or problem gambling jumped by about 30%, to nearly one in five.

In 2021, risky and problem gambling were much more common in men, young adults, Blacks, and people with household incomes in the lower and middle range - between $30,000 and $100,000 a year. (I’m guessing that luring that target demographic was the reason DraftKings chose Kevin Hart and LeBron James as pitchmen.)

The most important finding in the 2021 survey was that those who gambled online were more than 10 times as likely as those who gambled only in land-based venues to be problem gamblers.

The consequences

Of course, while people may win or lose on individual bets, in the long run every gambler loses. What happens when they gamble away big chunks of their income? A study that analyzed a massive trove of data from bank accounts showed that heavier gamblers skip credit card payments, take out payday loans, and miss payments on their car loans and mortgages.

Problem gamblers may also steal from relatives, friends and employers, and lose their jobs. Risky gamblers also spend less money on healthy and social activities.

All this harms their health, bigly. Problem gamblers are more likely to use alcohol or drugs and develop several types of mental illness, and they are much more likely to attempt or die by suicide.

Suicide is not the only fatal outcome, though. Gambling aside, people living on the financial edge have much shorter life expectancies and higher mortality rates from a long list of diseases, from heart disease to cancer. So when gambling companies vacuum up people’s money, the health effects are far-reaching. The bank account study showed a gradually increasing rise in overall mortality with increasing gambling. Over a five-year period, those who bet more than 30% of their monthly income were about 30% more likely to die.

The curve

The industry would have you think that gambling is only a problem for the small percent of their customers, who can be given individual help. By this line of argument, for others gambling is innocent pleasure. Here’s from DraftKings’ “commitment to promote responsible gaming”:

“While the vast majority of our customers play for fun and entertainment, we’re also committed to raising awareness of resources for those who may be facing problem gaming-related challenges.”

But as the graphs above show, the relationship between gambling and its harms is continuous. Sure, problem gamblers suffer a lot, but the much larger group of those at low or moderate risk are also damaged.

Treatment for gambling addiction can’t solve this. If the New Jersey survey is representative of the U.S. as a whole, this country has 15 million problem gamblers and 35 million at low to moderate risk. Only about 10% of problem gamblers seek treatment, and this treatment isn’t curative. That means treatment offers very little to nearly all of those 15 million problem gamblers – and nothing to the 35 million risky gamblers.

The solution

In 2025, DraftKings alone pulled in $6 billion in revenue. A study in Connecticut found that 70% of gambling revenue came from just 7% of the population who are risky and problem gamblers. That means that this one rich company has taken roughly $4 billion from Americans – mostly lower-income, many Black – whose gambling habits they can’t control and who can’t afford to lose that money. To me, this isn’t commerce. It’s theft.

In a saner time, our government would fix this injustice. I like the simplest solution: prohibiting online gambling. That would just bring us back to 2018.

But if that’s too radical an idea today, there are many ways to make online gambling less rapacious. For example, the government could restrict advertising by gambling companies, prohibit in-game bets and teaser promotions, limit the size and frequency of bets, and lower the percent of the handle that companies keep.

In concept, the solution is easy. But it won’t happen until enough of us admit that we have a very big problem.

Share

Subscribe now

By themselves, their arguments won’t carry the day. That’s what agribusinesses’ $200 million in political donations are for. But the arguments do matter. The more persuasive the food companies are, the easier it is for legislators to sign on, for Fox News commentators and other paid influencers to parrot the talking points, and for voters to be swayed.

What are they saying, and in the national conversation about ultra-processed food (UPF), how should we respond?

The case

The food industry’s talking points appear on two of their websites (here and here), and in this statement that their trade association gave to food journalist Helena Bottemiller Evich.

“America has one of the safest and most highly regulated food systems in the world…We welcome fact-based conversations about nutrition, but safety is not up for debate. As consumers continue to seek a diverse selection of foods and beverages, the makers of America’s trusted household brands provide a wide variety of affordable products to choose from, along with access to the information consumers need to make informed choices. Companies adhere to the rigorous evidence-based safety standards and nutrition policy established by the FDA to deliver safe, affordable and convenient products that consumers depend on every day.”

(Highlighting mine.) Taking these talking points one at a time:

• Safety

The industry must think that their strongest argument is that ultra-processed food is “safe” because they mentioned it four times.

And it’s true that ultra-processed food (UPF) is safe from contamination with bacteria. That’s a good thing. Before the Pure Food and Drug Act of 1906, contaminated food was a major source of disease and death. In part because of this law, deaths from foodborne diseases are now rare (here and here), and are mostly due to mistakes in kitchens, not contaminated grocery items.

But 120 years later, we’re overdue to update what we mean by food safety. People eating UPF take in an astonishing 500 more calories per day than people eating food that is not ultra-processed, gain weight fast, and are much more likely to become obese and develop type 2 diabetes. Obesity is causing an estimated 483,000 excess deaths a year, which arguably makes it the biggest killer in America.

So calling ultra-processed foods “safe” is dead wrong.

• Nutrition

Amazingly, the food companies are claiming that UPFs are good for our nutrition.

Mainly, that’s because many UPFs (including Twinkies and Froot Loops) are “fortified” with vitamins and minerals. No question, vitamin fortification is good for us. Adding folic acid to flour prevents more than > 1,000 babies each year from being born with devastating neural tube defects.

But most vitamin deficiencies are relics of the past that afflicted only people with extremely narrow diets. In my entire medical career, I’ve never seen or heard of anyone with pellagra, beriberi, rickets, or scurvy.

Regardless, adding vitamins and minerals to foods does not lead them to be classified as ultra-processed. No health expert is recommending that we stop fortification, and no proposals to limit ultra-processing would interfere with fortification.

The food companies also argue that ultra-processing enables them to produce food for people on specialized diets, like gluten-free or vegan. But none of the proposed policies on UPF would ban gluten-free cookies or soy milk.

• Affordability

Ultra-processed food is cheap…excuse me, affordable. In the debates about UPF, we’ll hear a lot about that. And we’ll hear that UPF therefore helps fight food insecurity, and that regulating it would exacerbate health disparities. Playing the poverty and race cards is a tried-and-true technique of the food industry.

UPF is cheap because it can be produced in huge quantities in cost-efficient factories. It contains preservatives that give it an incredibly long shelf life, which lowers costs by cutting spoilage and reducing resupply runs.

But, as I’ve written before, healthy food is actually no less affordable than processed junk food.

And it does not help people to offer them cheap food that makes them sick. Instead, it imposes costs on them - costs like the suffering from illness, lost work time, and out-of-pocket medical costs of obesity and its complications (which run about $4,000 a year for a person with type 2 diabetes).

Effectively, “affordable” UPFs just push industry’s costs onto us. And the consumers who pay the most are minorities and people with low-incomes.

• Convenience

Yes, UPFs are extraordinarily convenient. With UPFs, you don’t have to refrigerate, chop, cook, mix, stir, or clean up afterwards. And the decades-long shelf life enables food companies to place ready-to-eat UPF within arm’s reach everywhere, from gas stations to school vending machines to hardware stores. If convenience is all we care about, the food industry wins hands down.

But the long shelf life that makes UPFs so convenient is unnerving. On YouTube you can watch a guy slicing into – and then eating – a 20-year-old Twinkie. It had the consistency of a teething biscuit, but it was mold-free and otherwise completely Twinkie-like. That sure doesn’t make me feel good about what it would do inside my body.

I think we should value health – which, after all, often means the difference between life and death - above convenience.

• Consumer choice

American consumers have a right to eat whatever they want. For the government to restrict their choices is un-American. We will hear plenty of this argument, which taps into our deeply-held values.

I heard this argument in New York City in the early 2000s, when the Board of Health was considering a ban on trans fats in restaurant foods. The “grease police” were taking away your choice to eat tasty fries.

Since then, trans fat has been banned nationwide, and I have yet to hear anyone demand that we put trans fat back in their food.

But to the general point, I have three responses. First, a typical supermarket has some 30,000 items on its shelves, and the food industry introduces 10,000-20,000 new products each year. We don’t need more “food products” for people to choose from. We just need food.

Second, when companies spend billions each year in advertising to persuade us to eat ultra-processed junk, are we really making independent choices?

And third, health experts are not proposing a ban on UPFs. If you’ve come to depend on Twinkies to give your life meaning, you can rest easy.

• Access to information

Consumers already have access to information that they can use to make their own choices, the food companies argue. If people choose UPF, it’s because they’ve decided that they’re ok with how healthy it is or is not.

Sure enough, food packages are loaded with nutrition information. But that information has become available only over the fierce resistance of the food industry. For example, the food companies fought hard to prevent listing “added sugars” on the nutrition facts panel.

And the information on labels is utterly confusing. Can anyone without a degree in nutrition and a calculator make sense of this?

Under pressure to provide information that is easier to find and more interpretable, the food industry is pushing its Facts Up Front labeling scheme, which looks like this.

But is that any better?

What policies are proposed?

Any argument about UPF must consider what people propose we do about them. Given the enormity of the epidemic of diet-related diseases, health experts’ proposals are modest. They suggest that we:

• Restrict some techniques of marketing and promotion, for example, prohibiting placement of UPFs in supermarket checkout lines and limiting advertising to kids

• Require labels on UPFs that are either just informational (“ultra-processed”) or warnings (stop signs)

• Run counter-advertising campaigns, along the lines of Tips from Former Smokers

• Prohibit UPFs in schools and in other government-operated programs

• Restrict the use of SNAP benefits to buy UPFs

• Tax the most dangerous UPFs, like sugar-sweetened beverages

Bottom line

Agribusinesses have invented “food products” that are cheap, convenient, addictive, and - over time, for many people - fatal. Today we finally have some momentum in our fight against them. That fight will mostly involve political muscle, but one of the fields of battle will be the debate on the internet. It’s time for people who care about health to engage.

Share

Subscribe now

Between 1842 and 1861, James and Jane Lynch buried seven children, five between their first and fourth birthdays.

Today we take for granted that children rarely die, but that’s new. In the 1850 census, families were asked who had died in the previous year and from what cause. The mortality rate for young children recorded in that census was roughly 2,500 per 100,000, which is almost certainly an undercount. Even accepting that number, though, means that American parents on average lost 10% of their children between their first and fifth birthdays. The monument doesn’t say how many Lynch children survived, but if we assume that Jane bore a dozen children, more than 50% of hers didn’t live to reach age 10.

A good bet would be that the Lynch children died from gastrointestinal and respiratory infections. I took statistics on deaths of young children from that census and grouped them based on how common and preventable they were:

About a third of these deaths were counted as having been caused by cholera, dysentery, and other gastrointestinal infections, and even that’s probably a big underestimate because it misses many vaguely-described causes like “fever” that I’ve grouped as “other infections”. Adding tuberculosis, influenza, other respiratory infections, and diseases that we now prevent with vaccines, at least 75% of child deaths were due to infectious diseases.

Beneath these fatal infections lay a pathogenic neighborhood environment. The 1850s were early days in the industrial revolution, when people had moved from farms into cities to be near factories, or in Washington DC to be near government jobs. In those new cities, families maintained their country habits, using maybe an outhouse in the back yard and a water source that was a shallow well or a nearby spring or stream. Those habits were mostly safe on isolated farms but highly dangerous in crowded cities.

If the Lynches lived near the cemetery, odds are good that they got their water from the Anacostia River. That river was contaminated with human feces from people living upstream (it still is after a hard rain), making it a great killer during the cholera epidemics that terrified Washington in the 1850s.

Another toxic feature of cities at that time was crowding within homes, which enabled the spread of respiratory bugs like the influenza virus and tuberculosis bacteria. The two types of infections compounded each other; respiratory infections were much more likely to kill children who were malnourished after repeated bouts of diarrhea.

Make America Healthy Again, circa 1850?

The Lynches couldn’t have known that their water may have been a source of cholera, because John Snow’s hypothesis about it wasn’t accepted even in England until the mid-1860s. But they might have understood that something about the city was killing their children. If so, you couldn’t blame them if they were to consider moving out of that pestilence to a farm in Maryland – that is, trying to go back in time.

Instead, though, cities changed. In 1864, Washington’s city leaders built an aqueduct to bring in clean water from far upstream on the Potomac River. In the early 1900s, for added safety, they began filtering and chlorinating that water. In roughly those same years, city leaders throughout the U.S. built similar systems to supply uncontaminated water to their residents.

The clean water, combined with improved housing and nutrition, worked better than any could have imagined. From 1850 to 1950, mortality rates in U.S. children fell by at least 95%.

So while the Lynches surely endured many dark days, in retrospect they stood on the verge of what was probably the greatest improvement in child health in human history.

The Washington DC aqueduct was made possible by the prosperity caused by the industrial revolution, but it also required a new role for government: assuming public responsibility for what had been seen as a personal matter. As Geoffrey Rose put it, “The great public health reforms of the nineteenth century… were undertaken for people rather than by people.”

Lessons for today

Looking at today’s health problems, we blame features of today’s world, and with good reason. Our food system is driving epidemics of obesity and chronic diseases. The gun industry and the media marketing of violence are promoting shootings. Smartphones are causing depression.

In search of solutions, we tend to look to the past. It’s an alluring idea that this country was once an idyllic place that has become contaminated by modernity. That idea leads to the try to solve our problems by rejecting the modern world. That idea, with a fascist spin, is what animates Make America Great Again. Less hatefully, that idea is behind the Make America Healthy Again movement. The anti-junk food, anti-vaccine, anti-chemicals, and anti-GMO views fueling MAHA are fundamentally backward-looking.

A page on Robert Kennedy’s 2024 campaign website opened with this:

“The current generation is the sickest in American history. When John F. Kennedy was president, 6% of American kids had a chronic health condition. Today it is 60%.”

Of course, Kennedy’s numbers are complete bunk. But the quote’s outrageous lie is more dangerous than the made-up numbers. Children today are healthier than they’ve ever been. Mortality rates have fallen 75% since JFK was president.

What’s behind this recent fall in child mortality? Here are numbers for 1970 compared to last year:

Since 1970, deaths in children from infectious diseases, already a tiny fraction of what they were in 1850, have almost disappeared. Deaths from other causes – including injuries – have also plummeted. Give credit to vaccines, antibiotics, other forms of medical care, and regulation of the safety of products like cars and toys. That’s progress, not a fall from Eden.

Today’s solutions for today’s problems

Of course, we all want our children to be still healthier. While very few children die today, many are accumulating damage that will kill them later by eating ultra-processed food, being physical inactive, and watching video screens and smartphones.

The solutions to these 21st century problems do have some retro elements to them. Children should be eating more food that grows out of the ground and less that comes in a box. They should be outside playing with friends instead of inside watching screens. But our nation’s answer to these problems cannot be to return to the era of VW bugs and LSD. As in the 19^th century, our answer has to be taking public responsibility for protecting kids from our era’s major risks to their health. In the 19^th century that meant building an aqueduct. Today that means, for example, restricting the marketing of ultra-processed foods, regulating smartphones, and building the physical and social infrastructure for kids to play with each other.

Today – as in 1850 - the solutions to the health problems lie not behind but ahead of us. When we talk about making America’s children healthy, we have to stop saying “again”.

Share

Subscribe now

The swamp

New Orleans is a magical and tragic place. It is blessed with the nation’s best music – music that, even when it is sad, is joyful. And it is cursed by being governed by the State of Louisiana.

Louisiana’s government has a long tradition of incompetence and corruption, causing and caused by voters who are laughingly cynical. The voters elect politicians based on their entertainment value. Those politicians then line their pockets and appoint bumbling loyalists to run the government.

The tradition goes at least as far back as the early 1930s, with Huey Long, a governor who was as power-hungry as Trump and only slightly less self-obsessed.1 When told once that his actions were unconstitutional, Long said, “I’m the Constitution around here now.” One day, without authorization from the state legislature, Long ordered state prison inmates to demolish the Governor’s official home, which took all of six hours. He needed a grander mansion, he claimed, to host “visiting potentates and dignitaries.”

Later, Huey’s brother Earl Long served three kleptocratic terms as governor, one of which included a brief stay in a psychiatric inpatient unit. Earl once said that some day people in Louisiana would elect “good government, and they won’t like it.”

Edwin Edwards, who served four terms as governor in the 1970s through the 1990s, was indicted twice before being convicted of racketeering and spending eight years in federal prison. But he was great with one-liners. “The only way I can lose this election,” he once boasted, “is if I’m caught in bed with either a dead girl or a live boy.” Many voters loved him. “He’s crooked but he’s an honest crook,” said one.

When it came to appointing people to run state agencies, these and other Louisiana governors chose people based on their willingness to do what they were told. That required people with integrity scores of zero but proven incompetence.

The appointees didn’t have much time to get familiar with their own departments, because internal squabbling and their own mistakes caused many to be quickly ousted. With that, and with the tendency of incompetent managers to fire their subordinates on whim, the average tenure for state health officers in Louisiana was about 18 months.

Public health in the swamp

All that made it miserable – and risky - to be a civil servant trying to protect and promote health while answering to these political appointees. One very honest, capable person that I know was demoted and transferred to bureaucratic Siberia for refusing to sign a multi-million-dollar contract to her supervisor’s political friend for services that were, let’s say, ill-defined.

Layers further down, though, there were many career staff in the health department who were intelligent, dedicated, and competent - folks trying to fight epidemics and promote health while keeping their heads low. But it was tough for these people to break new ground, because innovation would inevitably annoy some constituency, which would irritate the boss. What appointee wants to fight political battles when there’s money to be made?

In the HIV/AIDS program, we tried innovating a bit. To promote safer sex, we attempted to flood the state with condoms, giving cartons of them for free to any organization willing to place condoms in bowls in gay bars, liquor stores, and storefronts in strips traversed by streetwalkers. One day a fundamentalist provocateur, claiming to represent a community nonprofit organization, got a stack of boxes of condoms from us, piled them up on his front lawn, and called in a local TV news crew. The state, he said to the camera, was passing out “toxic condoms.”

I tried and failed to get state legislators to even introduce a bill that would legalize syringe exchange. After that, I attempted to sneak funding for an underground syringe exchange program into a “community outreach” contract with a small community-based organization. The contract got snagged by eagle-eyed staff. I wasn’t fired for that transgression, but I sensed that I might be treading on thin ice.

So we were left mostly playing defense, that is, trying to block actions that would make the HIV/AIDS epidemic worse or were just generally evil. There were plenty of those to fight off. Every year legislators introduced bills that were harsh and counterproductive. Mandatory HIV testing and quarantine. Criminalizing spitting on another person. Blocking the distribution of condoms because, as one woman patiently explained to me, “all contraceptives are abortifacients.”

True innovation in public health happened elsewhere, with Louisiana following long afterward, if at all. For example, laws requiring restaurants and bars to be smoke-free started in California in the 1990s. It was only in 2006 that Louisiana got smoke-free restaurants, and the state still does not have smoke-free bars.

Playing defense in public health wasn’t the slightest bit satisfying. Who could feel good, after sidetracking a dangerous initiative, when victory meant that nothing would change? But blocking the worst actions saved lives.

Louisiana R Us

In those years, I hoped that Louisiana’s government might one day rise to the standard of the federal government – lumbering and bureaucratic, but basically honest and competent. Instead, the U.S. government has become like Louisiana’s, with corruption at the top and extreme loyalty and incompetence in the political layers underneath. Feeding this on the outside is extreme public cynicism about government and a tendency to vote for candidates who are entertaining.

The staff at CDC, who once had by many organizational layers insulating them from political shear forces, are now facing what Louisiana’s public health workers faced. That means they are mostly stuck playing defense. And those of us outside of CDC who care about public health are forced to spend our time crying foul when CDC staff can’t. Together, we’re trying to defend the whole nation against repeated attacks on vaccines, the decapitation of the smoking prevention program, the corrupt promotion of beef, the marketing of quack cures and “wellness” products, and the blundering and cruel response to the Ebola outbreak.

Other countries are now leading on public health policy. In facing down the epidemic of obesity and other diet-related diseases, the true innovation is happening in Latin America. Mexico passed a nationwide tax on sugar-sweetened beverages in 2014. Chile has required warning labels on the front of packages of unhealthy food since 2016. Brazil is already limiting ultra-processed foods in schools.

Of course, we all wish we could play offense. Who in public health doesn’t want to have the satisfaction of quelling epidemics or fighting off the biggest health problems of our time?

But defense is also crucial. In football, there is the same relative point value in preventing a touchdown as in scoring one. Likewise, lives saved by preventing scoundrels from killing people are just as valuable as lives saved by new technologies or policies. And if we must be followers rather than leaders in public health – as with the diet-related disease epidemic – then let’s follow.

So to those beleaguered folks in CDC or on the outside trying to help: keep at it, even if you have to rethink what counts as progress.2

There are no prizes awarded for public health workers who save lives by playing great defense. But if there were, I know a few that I’d like to nominate.

Share

Subscribe now

When health risks fade, those of us who promote health tend to claim credit, and when they get worse, we tend to blame forces outside of our control. This alcohol trend makes me very uneasy, though. Other than some mostly-ineffective presentations in high school health classes, this country has almost no organized efforts to reduce drinking. So why would young people put aside booze?

Trends

People in Gen Z, born between 1997 and 2011, are now 15-29 years old. The story that they are drinking less is not just an urban legend, and the change is not small.

In the National Survey on Drug Use and Health (NSDUH), young adults aged 18-25 have been steadily drinking less since about 2010. In teens aged 12-17, the drop-off has been going on for much longer - at least the past 30 years – and is even more impressive.1 Teens in 2024 were less than half as likely to drink as their parents in 1994.

In the Monitoring the Future project, which has been surveying middle- and high-schoolers for decades, it looks like the fall in teen drinking started way back in mid-1980s.

And this isn’t just a U.S. thing. Teens and young adults are also drinking less in Europe, Australia and New Zealand.

Theories

Plenty of theories have been floated (here and here) to explain this, but none is widely accepted. Taking them one at a time…

Cost?

One argument that I’ll dispense with quickly is that Gen Z can’t afford to buy booze because it’s broke.

The argument doesn’t pass the sniff test. One or two teens can get drunk on a bottle of cheap wine for less than $5. Many families are financially squeezed these days, but I can’t believe that most young people in America looking to get high can’t find that much. Also, this theory wouldn’t explain a steady decline in drinking through booms and busts of the economy over the past 30+ years.

Changing drug of choice?

Could young people just be switching to marijuana to get high? That’s a tempting explanation. In NSDUH, among young adults, the drop in drinking has been mirrored by a big increase in marijuana use:

But the pattern doesn’t hold for teens, who since 2000 have been cutting back on both drugs. Here are trends from CDC’s Youth Risk Behavior Survey:

Health consciousness?

Have Gen Z’s just finally gotten the message that alcohol is dangerous? I wish that were true, but that theory doesn’t stand up to scrutiny.

NSDUH tracks how risky people believe different drugs are. Here are trends in the percent of people who see drinking five or more drinks once or twice a week as a “great risk”. While this risk perception has been creeping up a bit since about 2010, it’s much lower than it was in the 1990s.

A related theory is that young people are just becoming hesitant to take any risk. This theory is appealing because young adults today also less likely to have sex:

And, since the 1990s, high school students have become much less likely to do a bunch of unrelated risky things, like have sex, smoke, use cocaine, or drive unsafely.

But this unsatisfying explanation just prompts a second question. Why would teens and young adults have become risk-averse?

Helicopter parenting?

Here’s a theory on that: the shrinking family size.

I’m a Boomer who grew up in a very large family, in a town where families with four or more kids were common. When I was a teen, my parents couldn’t – and didn’t even try – to track where I was or what I was doing. I wasn’t a big risk-taker, but I certainly did some stupid things that they never know about but wouldn’t have approved of.

Families are much smaller now. The average number of children in households with at least one child has fallen drastically in the past 170 years, hitting a low in 1990, about when teen drinking started to fall.

Fewer kids at home means parents can hover over the ones that they have. And we know parents very much don’t want their kids to drink. Every year in NSDUH, 90% of teens say that their parents would strongly disapprove of their drinking 1-2 drinks a day. Parents likewise don’t want their teens to smoke, use cocaine, have sex, or drive without a seatbelt.

By the time young people hit their early 20’s, they are generally more able to misbehave without their parents knowing, so the helicopter parent theory alone might not explain the drop in drinking in young adults. But studies that track people over years show that those who don’t drink as teens tend to retain their nondrinking habits when they become young adults.

Smartphones?

The downward trend in drinking in young adults accelerated in about 2010. That was right around the time that smartphones took off. I can think of two reasons why smartphones might cause teens and young adults to drink less. First, it means that parents can keep in even closer contact with their kids - even those in their 20s.

Second – and I think more important - smartphones have driven a shocking rise in social isolation. Young people are now scrolling Instagram or watching TikTok videos instead of hanging out with friends. In the American Time Use Survey, since 2010 the time that teens and young adults (in the blue lines below) spent alone rose by about 40% and the time that they spent with friends plummeted by two-thirds.

Teens and young adults typically drink only when they’re relaxing or partying with friends. Less time with friends means less drinking.

This explanation aligns with the smartphone theory of everything, which blames nearly every social disruption over the past 15 years on those attention-grabbing devices. Others have pointed out that smartphones are making young people lonely and less likely to have less sex. And, as I’ve written about before, smartphone use appears to be driving increases in depression, suicidal thinking, attempted suicides, and completed suicides, not just in teens but also in adults.

The big drop in drinking, then, may be just a thin silver lining to a very dark cloud.

Bottom line

Bottom line: we shouldn’t be declaring victory for Gen Z’s turn away from alcohol. They may just be giving up drunken partying for doomscrolling.

This is one more reason that young people – and the rest of us – need to figure out how to put down the phones and spend more time with other humans.

Subscribe now

Share

As someone very worried about our democracy, I’m relieved that Fetterman isn’t converting to Republicanism. But I’m not reassured.

It’s not Fetterman’s endorsement of steak that bothers me. It’s that a surefire way to demonstrate your Democratic bona-fides these days is to be pro-weed.

Fetterman is not new to his position, and there’s no question that he means what he says. Marijuana legalization has been his signature issue since he was Pennsylvania’s Lieutenant Governor. As far as I could tell, the only thing he did in that office was hold community meetings about cannabis legalization in every one of the state’s 67 counties.

With all the problems that our nation faces today, marijuana legalization hardly seems the right hill to die on. Especially because the drug’s literally driving people crazy.

I’ve written about this before, but the evidence is now stronger than ever that we are watching a multi-national epidemic of psychosis caused by marijuana, and I feel obligated to keep ringing alarm bells.

Weed trends

Medical marijuana is now legal in 40 states and recreational marijuana is legal in 24. Several other states may legalize recreational cannabis in November’s election, and the federal government is now on track to loosen the regulations on it. I’m guessing that full federal legalization will happen by 2030.

In the states that have already legalized recreational weed, marijuana-related arrests have plummeted by nearly 90%. That sharp drop in arrests counts as a victory among those (myself included) who see these arrests as destructive government over-reach. But with the dropping of laws, use of marijuana is rising fast, with early legalization states leading the way.

Cannabis and mental illness

The consequence that I worry about most is the damage to users’ brains. Cannabis use has been linked to many types of mental illness, including addiction (politely called cannabis use disorder or CUD), suicide, schizophrenia and other forms of psychosis.

People with psychosis have a deeply distorted view of the world, experiencing hallucinations and delusions that range from unpleasant to highly dangerous. The risk of psychosis rises sharply with more potent forms of the drug, and the potency of marijuana has skyrocketed.

With high potency and increasing use, I would expect to see increases in the number of people diagnosed with psychosis in the U.S. Unfortunately, while some local data do show increases in psychosis, there is no national data system to track it.

Other countries are in a better position to measure this because they have medical registries that cover entire populations. Recent studies from places that have these registries are showing that the rise in marijuana-related psychosis is big.

Denmark

In Denmark, medical marijuana use is legal and recreational use is illegal but tolerated. With huge increases in cannabis potency there, the number of Danes diagnosed with cannabis use disorder surged 10-fold between 1990 and 2016.

And over the same years, the incidence of schizophrenia more than doubled.

And the percent of cases of schizophrenia that can be attributed to CUD quadrupled.

Recently, Danish researchers looked at new diagnoses of psychosis by birth cohort. Men born after 2000 were about 30% more likely than those born in the 1980s to be diagnosed with schizophrenia or psychosis by their early 20s. For women, rates by age 20 nearly doubled.

Canada

Canada legalized medical marijuana in 2014 and recreational marijuana in 2018. Like Denmark, Ontario saw a huge rise in cannabis use disorder from 2006 to 2022.

And that province saw a doubling of “psychosis not otherwise specified” over the same years.

Also, as in Denmark, people in Ontario are increasingly getting diagnosed with psychosis at younger ages. Those born after 2000 were about twice as likely to have psychosis by age 20 than those born in the 1970s.

Stemming an epidemic

It’s painful to watch this epidemic unfold. People with schizophrenia and other forms of psychosis really suffer, and their family members struggle mightily to keep their loved ones safe and functional. Our approach to mental illness in this country is entirely focused on helping people who are already ill get treatment. Here’s a rare opportunity to prevent the most serious form of mental illness – an opportunity that we are missing.

The marijuana legalization push comes from users, social justice activists, legislators hungry for tax revenue, and – especially - a marijuana industry hungry for profits. Fetterman’s pro-legalization argument echoes that of social justice advocates: arresting users causes them lifelong damage (giving them job-destroying criminal records) and prohibition merely creates a toxic black market.

I don’t disagree. But psychosis can be a life-destroying problem, too.

We don’t have to choose one awful consequence or the other, though. We can have legalization without commercialization. That is, we can legalize the use of marijuana, but regulate the sales and elimination promotion in ways that protect Americans’ health, as I described in an oped a while back. For example, we could ban advertising and other forms of marketing, make the drug available only in state-run outlets (as some states currently operate for wine and spirits), and regulate the product to reduce potency to much-safer levels.

Please, someone tell Fetterman and the rest of the Democrats that.

Share

Subscribe now

I posted this on my Substack site a few days ago, and it seemed to strike a chord with the few people who saw it there, so I’m now sending it to everyone. Drop in a comment to let me know what you think.

I don’t use AI to write.

Here’s what I mean by that: I don’t pose questions to a large language model (LLM) and then post the responses. I don’t take the AI output as my first draft and then edit it. I don’t use AI to “just organize my thoughts”. I don’t cut-and-paste text from AI into my articles. I don’t write a first draft and then run it through AI for “light editing”.

Now, I do use AI to find sources of information. But I always then pull the original sources and decide for myself what they are saying, rather than use AI’s summaries. That’s because AI’s summaries, in my experience, are usually wrong. Sometimes they are only slightly off on unimportant details, but sometimes they are dead wrong on crucial details.

Why do I feel compelled to say this? Because it seems that everyone else is using AI to write. Publishers are now flooded with books written by AI – some proudly, some surreptitiously. In classes that I teach, despite being given clear rules to the contrary, some students use AI to write their assignments. Other Substack writers post articles that, based on the syntax and content, must have been written by AI. And even a hot new book exploring the nature of truth in the AI age contained quotes made up by AI.

So now when I read others’ posts, I’m not sure who’s speaking. Is it the guy whose name is below the title, or Claude? I find it creepy.

Why not?

If using AI to write is so common, what’s wrong with it? For me, three reasons.

First, if I’m using AI to write, you don’t need me. Anyone can ask ChatGPT a question. If that’s all I’m doing, I’ve just proven that I’m useless.

Second, AI offers only the conventional wisdom. To my (limited) understanding of how LLMs work, they just “read” and synthesize the entire world’s literature. What they produce is akin to the average thought of everyone who writes or speaks into the internet. That represents one heck of a lot of knowledge. I’d estimate that the conventional wisdom that comes out of LLMs is (in broad strokes) correct maybe 90% of the time, which is a much better batting average than we humans have.

But when I read, I want to learn the unconventional wisdom, too. People drawing conclusions that others haven’t is how knowledge grows. Before Darwin, the conventional wisdom was that all species were dropped on Earth by a Supreme Being, and afterward no species were either created or destroyed. Evolution – today’s central organizing principle of biology - was thought up by a man who made his own observations and then ignored the conventional explanations.

In intelligence and creativity, I’m a very long way from Darwin. But I at least try to think originally – to reason from facts and evidence, rather than just accepting what others have said. Because they don’t reason, LLMs can’t do that.

I suppose I could just use AI for “light editing” or “first drafts only”. And I’m guessing that the future’s best writers will do something like that, combining their human brains’ unique abilities with the awesome strengths of AI, just as the future’s best doctors will combine their own wise judgement with the awesome capability of AI to keep up with – and synthesize – the entirety of human knowledge of diseases and treatments.

Maybe in the future, writing without AI will become a quaint niche craft, like basket-weaving. But if I start using AI to shoulder part of the load of writing, where would I draw the line? Writing is hard. It is so much easier to have a machine do it for me. With that gravitational pull, I suspect I would slide down the slippery slope to assigning AI the entire job, even if I cling to the delusion that I’m “writing”.

Third, and most important, writing is thinking. There is nothing like staring at a blank screen, struggling to produce a sentence, to force me to think hard. What do I really understand and believe about this topic, and why? Am I sure? If I let AI write for me, eventually I will lose the ability to think, just as a person who drives everywhere eventually loses the ability to walk to places. Someday, if I live long enough, I will lose all my marbles, but I refuse to just give them away today.

And there is a coda to this last reason. Today, AI-generated text is free and, as far as I can tell, absent of paid promotions. But when AI investors start demanding a return on their multi-billion-dollar investments, LLMs will change. Chatbots will be available only by paid subscription, or stuffed with advertisements, or (worse still) full of embedded product placements. How will it feel when we can no longer write without these for-profit bots?

So what?

Here are the consequences of my no-AI pledge:

I will never be able to post every day. I write slowly and with plenty of time-wasting inner turmoil. There are some intelligent writers (like Paul Krugman and Noah Smith) who are stunningly prolific. I’m not one of them.

I make human errors. Typos, missing words, mangled grammar. Even with word processors’ spellchecking and grammar hints, some mistakes slip through. As John McPhee wrote, “Typographical errors are more elusive than cougars.” LLMs don’t make typos.

I generate sentences that are awkward. Re-reading old posts, I almost always find passages that make me cringe. Maybe these days I should wear this clumsiness as a badge of pride – evidence of my humanness – but I’m embarrassed anyway.

Still, you know when you read one of my posts that you’re learning what I - a real, flawed human - think.

Share

Subscribe now

The most important element of MAHA, according to survey respondents, was food policy.

It’s shocking that, rhetorically at least, Trumpist Republicans are leading the charge here. Robert Kennedy, Jr. has called ultra-processed food “poison” and promised to “change regulations, research topics, and subsidies to reduce the dominance of ultra-processed food.” His MAHA assessment report devoted 16 pages to ultra-processed food, and he’s encouraged governors to ban ultra-processed foods from schools.

Last year, Secretary of Agriculture Brooke Rollins, describing herself as a MAHA mom, welcomed a pack of other MAHA moms to lobby Washington for “the righteous work to ensure that we have the best, the safest, and the most transparent food supply in American history.”

Democrats, who have traditionally led on health, were caught off guard, but now see MAHA as a political opening. Chellie Pingree, a Democratic Representative from Maine who is also an organic farmer, told Time magazine, “Don’t miss this opportunity. Let’s find areas where we can find a win, like healthy food or [pesticides].”

Now, a group of food experts have handed them the tools to do that.

Ultra-processed foods vs. health

MAHA moms are right in seeing modern food as a major health threat. Estimates by the Global Burden of Disease project are that unhealthy diets kill some 350,000 Americans a year, leaving junk food a close second to tobacco in the rankings of leading underlying causes of death.

Of the many ways to categorize food that is unhealthy, the “ultra-processed” label has caught on in recent years, not just among politicians but also among health experts. Strong evidence links ultra-processed food to the most important health problems that we face, including obesity, diabetes, heart disease, anxiety, and depression. No one is entirely sure why, but several mechanisms are likely at play. Ultra-processed food is stripped of its natural fiber and “food matrix”, packs many calories per gram, and is full of the additives that stimulate our brain’s pleasure centers but trash our arteries – salt, sugar, and saturated fat.

Today more than 50% of our calories come from ultra-processed foods, a percentage that is growing, not just here by also around the world.

But anyone trying to turn that trend around has faced what may seem like a surprising stumbling block: exactly what counts as ultra-processed. We can’t rid our diets of this stuff if we can’t define the term precisely.

In search of a practical definition

The ultra-processed concept came from Brazilian epidemiologist and social activist Carlos Monteiro. His definition runs into the hundreds of words and is filled with confusing words and vague terms like “mostly”, “often”, and “typically.” Turning that concept into a clear, usable definition that food companies can’t game has proven to be hard, even for health experts. A food label tells you the ingredients, not the processes that the food manufacturer used to make it.

Some state legislators have introduced laws on ultra-processed foods (UPFs), and most of those laws define these foods based on specific artificial ingredients, like food dyes, emulsifiers, or preservatives. These definitions end up being very narrow, covering only about 10% of new packaged food products, compared to about 75% in Monteiro’s definition.

To fulfill his campaign promises, Kennedy needed to first create a UPF definition for use by the U.S. government. That also was a convenient excuse for inaction. Nearly a year ago, Kennedy posted a call for input on the definition. Last September, his MAHA Strategy report made much of the task of defining UPFs (and did little else). This past February, Kennedy told Joe Rogan that he would publish the definition “by April”. Last month, he told Congress that the UPF definition had been drafted and was being reviewed by non-HHS agencies.

Count me as skeptical that Kennedy will produce any definition soon, or a useful one ever. Food companies hate the idea that anything they produce would earn the poison label. They’re pushing back hard.

If the feds do come up with a definition, the food industry wants it to be as narrow as possible. It’s hard to imagine that Twinkies would escape the designation, but food companies are deeply worried about their yogurt, peanut butter, bread, and cereals.

Now we have one, and more

Now a group of the nation’s most accomplished food experts has laid down a marker. The group included food policy experts, researchers in nutrition and in the promotion of healthy eating, representatives of consumer groups, and Monteiro himself. They came up with a definition that isn’t elegant, but is workable:

• A food is ultra-processed if its ingredients include either a “cosmetic additive” (like flavors, colors, thickeners, or emulsifier) or a “non-culinary ingredient” (like high-fructose corn syrup, lecithin or interesterified oils) other than vitamins, minerals, herbs, spices, and yeast.

The group excluded from the definition food items that the FDA already classifies as healthy based on their having adequate amounts of recommended food groups, no artificial sweeteners, and low amounts of salt, sugar, and saturated fat.

To make the UPF definition work, the FDA would have to create and maintain a list of definition-earning cosmetic additives and non-culinary ingredients. That wouldn’t be hard.

This is a broad definition that sorts food much like Monteiro’s original. The group’s reasoning for going broad was that the studies showing that UPFs are unhealthy used Monteiro’s definition, and there is no scientific justification for narrowing it.

That definition alone counts as progress, but the experts then did something else incredibly important. They reviewed a bunch of policy ideas that might help Americans cut their UPF intake and scored the ideas on their potential impact and feasibility. The policies that they scored highest were:

• Changing the Dietary Guidelines for Americans to recommend that people avoid UPFs

• Requiring UPFs to have front-of-package labels saying so

• Eliminating or reducing UPFs served in schools and purchased in other government programs

• Taxing the worst UPFs, like sweetened beverages

• Running media campaigns warning people about UPFs

I like many of the other ideas, too, but this list would be a great start.

Your move, MAHA

With this, Trumpist Republicans no longer have excuses to stall on UPFs. They’ve got not just a workable definition, but also ready-made policies that would protect Americans from nearly our greatest health risk. If they truly want to make America healthy, all that Kennedy and the Republicans in Congress need to do is say yes.

Of course, I don’t think for a minute that they will do that. Republicans are way too much under the thumb of agribusinesses and the big food companies. Case in point: Trump’s order to increase the production of the hated herbicide glyphosate, which was shamefully endorsed by Kennedy.

Democrats in Congress probably aren’t much better. But we’re approaching pivotal mid-term elections, with the race for the presidency soon to follow. I’m no political consultant, but it sure seems to me that a policy with 78% public approval is one that candidates should run on.

So – speaking to MAHA moms now – whether you vote for Republicans or Democrats, if you truly care about protecting your child’s health, you should hand this report to candidates and demand that they sign on.

Subscribe now

Share

The article was written by Times regular Zeynep Tufekci, described as “a computer programmer by profession and academic training before turning her focus to the impact of technology on society and social change.” I’m a doctor, trained in epidemiology by CDC, who spent years investigating disease outbreaks. What could she possibly say about a disease outbreak that’s worth my time?

But after reading her piece and digging a bit, I realize that she’s right. We should take hantavirus more seriously. I only wish I could trust the Trump-ravaged health agencies to do it.

Tufekci points to key assumptions that the World Health Organization appears to be making that may not be true, about:

• How readily hantavirus can be spread. WHO is saying that the Andes strain of hantavirus requires “close and prolonged contact” to spread from one person to another. But in a 2018 outbreak, this virus managed to spread readily in various settings over four successive waves. It infected five people at a 90-minute birthday party, 10 at a wake, and two within hospitals.

In several instances, the contact between the infectious person and the infected person didn’t seem particularly close or prolonged. One person who became infected at the birthday party may simply have said hello to the infectious person on a walk to bathroom. One person infected at a hospital was an administrator who visited patients in their rooms but didn’t get close to them or touch them.

• The route that hantavirus takes. WHO is recommending that people who may be incubating the virus wash their hands frequently but not wear masks unless they have respiratory symptoms. The agency appears to be assuming that the virus spreads mostly through direct or indirect contact. But in the 2018 outbreak, investigators concluded that the virus was often spread through the air by “inhalation of droplets or aerosolized virions.”

• When people with the virus start spreading. The WHO’s guidance mostly assumes that people with hantavirus only start spreading the virus when they develop symptoms like a fever or a cough – though the agency does say “pre-symptomatic transmission cannot be entirely ruled out.” But one person in the 2018 outbreak spread the virus to six others “during the early prodromal phase” of the infection, which could mean when symptoms were too mild to be noticed. And the lead researcher believes that people may become infectious 48 hours before they develop symptoms. The timing here is crucial, because pre-symptomatic spread makes it impossible to isolate infectious people before they pass the virus on to others. It proved to be a major reason why contact tracing teams failed to stop COVID-19.

To be fair to WHO, most of the wording in its recommendations are cautious, like “current evidence points to…” and “as a precautionary principle…”. I believe its overall summary is accurate: “This is not another COVID. The current public health risk from hantavirus remains low”.

And I’m still guessing that this cluster will burn out quickly. The virus doesn’t seem nearly as easily spread as SARS-CoV-2. It’s somewhat reassuring that this outbreak took place on a cruise ship, a setting that is notoriously fertile ground for virus spread. Given the long (up to 40 days) incubation period of hantavirus, I wouldn’t be surprised to see a few more ship passengers becoming ill over the next couple of weeks, but with the attention that this outbreak is getting, I doubt they’ll pass the virus on to others.

Still, it is reasonable for those of us who work in public health to be a little edgy:

• We don’t know a lot about hantavirus yet. It was only discovered in the 1990s, infections are rare, and research on it is thin.

• The Andes strain is highly infectious, behaving differently from other strains, which don’t seem capable of jumping from human to human. WHO may be inappropriately applying lessons from the other strains to this one.

• The virus can evolve. In 2020-21, I was shocked by how quickly SARS-CoV-2 evolved into new strains (like Omicron) that were more easily spread than the original.

We need a little humility here. Too much is uncertain. In our attempts to pre-empt a counter-productive COVID-19-like panic, we shouldn’t offer reassurances that may not stand up over time.

Here’s the kicker, though: the uncertainty in this outbreak cries out for a top-notch gang of virus experts. We need epidemiologists and lab scientists at CDC studying the questions that Tufekci posed. We need researchers at NIH and universities studying how this strain affects human cells and what its vulnerabilities are. And we need vaccine developers in pharma turning the virus’ RNA sequence into an mRNA vaccine, just in case. We had all those folks in place until Elon Musk and Robert Kennedy, Jr. did their best to run them off.

Share

Subscribe now

People who work in public health, who typically lean liberal, usually side with those pushing for more medical care. It’s front and center on the American Public Health Association’s website: “APHA works to improve access to care…”

I’m here to argue that we need to cut spending on medical care – not to cut taxes (although it might help do that) or to benefit our economy (although it might do that) or to help people struggling with out-of-pocket costs (although it should do that) – but instead because it would improve our health.

The costs

Medical costs are crushingly high and still growing. In 2024, Americans spent $15,000 per person for medical care. Even adjusting for inflation, that’s more than 7 times what we spent in 1970.

Other wealthy countries spend on average about half that.

Where do those hefty payments go? Spending for every type of medical care has fast outpaced inflation, but the biggest cost driver is hospital care, which has roughly doubled in just the past 20 years. (The increase is certainly not being driven by primary care, which amounts to only 5-8% of total medical care spending.)

Flush with all that money, hospitals have become the cathedrals of our time, with soaring atriums filled with expensive art. Maybe some patients see that opulence as a sign that they’ll get quality medical care. I see it and think, “I’m paying for this.”

Cost shifting

Reinstating the enhanced Obamacare subsidies would lower some people’s out-of-pocket costs and persuade many to keep paying their health insurance premiums. But that would do nothing to cut our overall health care spending. It’s not cost reduction, it’s cost shifting.

And that’s the point that people on the liberal side tend to overlook. Ultimately, all medical costs come out of our pockets. We pay those costs in: our own insurance premiums; our employer’s contribution to insurance premiums (which could otherwise go to higher wages); higher costs of goods and services so that other employers can pay those premiums; and taxes for Medicare and Medicaid.

Those costs and taxes, by the way, tend to be regressive – that is, poor people pay a higher percentage of their income for them than rich people. Jeff Bezos isn’t paying those exorbitant hospital charges. Ordinary Americans are.

And those inflated costs are harming ordinary Americans. Imagine if the U.S. were to lower its per-capita medical care spending to that of, say, Germany. That would save Americans on average $6,000 per person per year, while still giving everyone top-notch care. To put it another way, our nightmare-to-navigate medical care system is charging a family of four an additional premium of $24,000 a year. No wonder families are having trouble paying the rent.

Diversion

Around the world, governments must make tough decisions on how to allocate their tax revenue. How much should go to education, roads, pensions, medical care, policing, weapons?

Among high-income countries, the United States is an outlier in those allocations. The colossal spending on medical care is paired with low spending for social services, like unemployment benefits and family support payments. Put medical and social spending together, though, and the U.S. is in line with other rich countries. Effectively, our government leaders have decided to direct money that might have gone to social services into medical care.

And that tradeoff actually happens. I’ve seen it play out each year in the budget fights in Louisiana, where I lived for a long time. Spending for Medicaid, an entitlement, is out of legislators’ control. But spending for education or social services is discretionary; each year legislators decide how much to allocate for them. So when Medicaid costs rise another 10%, unless tax revenue grows in parallel, spending for education and social services must be cut, even in a state with nearly the lowest ranking in education in the nation.

Social determinants of health

Why is that a problem for health? Because the social conditions in which people live have a powerful impact on how healthy they are. People who complete a high school education have a mortality rate about 25% lower than those who don’t. Even education in early childhood is associated with a 20% reduction in cardiovascular disease. A childhood spent in poverty leads a wide range of diseases in adult years. Having unstable housing is a big risk to health, and providing housing support has been associated with improvements in health.

It shouldn’t be surprising, then, that studies comparing different countries or different states in the U.S. have found that spending for social services is more closely linked to health than spending for medical care is. For the sake of our nation’s health, it’s smarter to spend $1 million educating poor kids than to buy a few more hospital stays.

Use health insurance?

Today, plenty of experts are arguing that to promote health we should provide social services, like housing support. And programs in which case workers help people deal with various social needs seem effective, that is, they reduce visits to hospital emergency rooms and the costs of medical care (here and here). But proposals on how to pay for those services have taken a very strange turn: bill health insurance.

That’s an incredibly cumbersome and inefficient way to do it. The health insurance system is designed to pay providers to give services to specific patients, with each provider submitting an insurance claim for each service. Imagine having to submit a claim to a health insurance company every time a teacher helps a child in a school. And imagine not being able to submit claims for the costs of that school’s principal, guidance counselors, janitors, or cafeteria workers.

Social services like education and housing are best delivered to groups of people, with the organization providing those services paid through a contract. If we want to promote health, governments should provide education and housing, not run the charges through the medical care system.

How do we cut costs?

The American medical care system costs way too much because it is rife with perverse incentives to spend more. Doctors get more money by doing more tests and procedures. Hospitals get more money by having more hospitalizations. Insurance companies get more money by paying more claims, from which they skim a percentage. And all those players make more money by charging higher prices, with no one on the receiving end in a position to negotiate. To cut costs, rather than just shifting them, we need to change those incentives, while still structuring the system so that people have access to curative and preventive care.

I’m not an expert in this, but I see two options. One is to change to a single payer system, like Canada’s, and use that leverage to reduce payments. The other is to fix the incentives in the existing system. Ashish Jha has just posted a wonderful multi-part series on how to do this, with ideas like paying providers to keep people healthy instead of treating sickness, site-neutral hospital payments, and prevention of hospital monopolies.

Ultimately, the obstacles to cutting medical spending are not theoretical or practical. The obstacles are only political – fierce resistance from the players in today’s system that don’t want to lose their lucrative positions. Overcoming those obstacles will take a political push from all of us, including those who care most about health.

Share

Subscribe now

Reminder: alcohol is toxic

Alcohol is by far the biggest public health problem that we aren’t even trying to solve. It kills, by CDC estimates, some 178,000 Americans a year. Those deaths show up as injuries (suicides, car crashes, and mixed alcohol/drug overdoses), chronic damage to the liver and pancreas, heart disease, stroke, and cancer. The victims aren’t just drinkers. Plenty of families injured by drunk drivers, adults shot by drunk friends, and children beaten by drunk parents can attest.

CDC and state health departments have (or had until recently) offices that combat smoking, educate people about nutrition and promote physical activity. But public efforts to reduce those 178,000 alcohol-caused deaths are almost nonexistent; government-funded alcoholism treatment programs barely scratch the surface and a tiny prevention program at CDC can do little. In our national conversation on alcohol, then, the companies selling the drug are the only voices speaking.

And those companies are speaking plenty, with billions of dollars of ads designed to persuade us that drinking will make us happier, more popular, sexier, more masculine, or whatever else we crave.

The Prohibition era taught us that some people will always drink and that people should be legally allowed to drink. But do we really want alcohol companies encouraging people to take this toxic drug?

Ads sell the product

Alcohol advertising works at two levels. It persuades individuals to drink (here and here). And it builds a culture that supports drinking (for example, here) by convincing people that a group can’t have fun without booze. Can you imagine a crowd cheering at a football or baseball game without beer? If not, it’s because a lifetime of seeing alcohol ads has forged a link to them in your brain.

One telling example of the power of these ads: in 1996, after a decades-long quasi-voluntary ban, liquor companies began advertising distilled spirits on radio and television. Here’s what happened to sales afterward:

What about restrictions?

Taking cigarette ads off the air in 1971 proved to be a crucial step in the slow but largely successful battle to reduce smoking. We’re so saturated with booze ads that today people have trouble imagining a world without them, but getting rid of them could similarly open the gateway to finally reducing alcohol’s damage. That’s why the World Health Organization recommends bans or tight restrictions on alcohol advertising, sponsorship and promotion.

In the U.S., support for alcohol ad restrictions today is soft at best. In a recent survey, about half of drinkers favored prohibiting broadcast alcohol ads when children are likely watching, but only 20% supported prohibiting booze ads late at night.

But other countries are making (admittedly weak) attempts. For example, in Brazil distilled spirits can’t be advertised on broadcasts between 6 AM and 9 PM, and in Mexico, alcohol ads can’t associate drinking with to success, popularity, attractiveness, or social acceptance.

Ads go digital

But even as countries take these tentative steps, the alcohol companies are shifting their ad dollars from TV, radio, and print channels to digital. That includes banner ads on websites, but banners are fairly easy to ignore. A more targeted and successful approach is embedding ads into social media feeds. By hitching onto these distribution channels, alcohol marketers can direct ads to specific subgroups incredibly cheaply. For example, researchers in Australia found that they could deliver alcohol ads to teens via Facebook for only $2.00 per 1,000. These digital ads work – that is, they increase kids’ drinking.

The new study tracked alcohol ads delivered via social media in Brazil, Mexico, the Philippines, and South Africa. The laws designed to restrict alcohol ads in these countries were written with traditional television and radio in mind, and they barely address these kinds of ads. Over a single month, the researchers tallied about 4,000 social media posts advertising beer, wine, distilled spirits, and “alcopops”. About ¾ of these ads appeared on Facebook, Instagram, and X. These posts generated an estimated 2 billion impressions, or about 1.5 exposures per internet user in those countries. Here are some examples:

From Brazil:

From the Philippines:

Many were directed to young people and to women, like this one from the Philippines:

So, even in countries on the other side of the earth, Meta and X are making money by persuading young people to drink.

Now comes artificial intelligence

One major advantage that Meta has in selling advertising on Facebook and Instagram is that it has deep information about its users – not just their age, race, and sex, but also their interests. It sells use of those profiles to advertisers. For example, a company can deliver alcohol ads specifically to kids age 13-17 whose online activity shows that they are interested in drinking. The rest of us – including those kids’ parents - wouldn’t know anything about the ads because we don’t see them.

That’s bad enough, but AI can take targeting of ads to the next level. People already tell chatbots their deepest secrets – the fears and desires that they don’t tell their closest friends or spouses. The alcohol industry can embed itself in these chats, shaping their messages to users’ particular psychological weaknesses. It’s not hard to imagine a chatbot burbling along, helping a young woman cope with a particularly stressful day, suggesting that she might feel better if she were to have a drink.

Think that won’t happen? Alcohol companies are already using AI to create new products and to advertise specific products to match people’s taste preferences. Alcohol-promoting chatbots would be an irresistible next step that offers profits for both alcohol companies and AI companies. After all, somebody has to pay for those expensive AI data centers.

P(doom)?

AI theorists talk about P(doom) – the probability that AI will cause an existential catastrophe, like a bioweapon that kills all of us humans. My own guess is that the odds of that are low, in part because so many people are warning about it. On the other hand, without strong government-imposed limits, I’d estimate the probability that AI will manipulate people into using alcohol (and maybe other dangerous drugs) at 1.0. That scenario won’t wipe humans off the face of the earth, but it certainly will make plenty of people die or suffer unnecessarily. So shouldn’t we be talking more about that?

Share

Subscribe now