Childhood Roundup #2

Recently: The Kids are Not Okay, Childhood Roundup #1, Fertility Roundup #1

It seems like a good time to go over the interesting relevant stuff I’ve found since the last time I did one of these.

Executive Summary and Table of Contents:

1. Why Are the Kids are Not Okay. Additional fleshing out of the ‘school as giant hypocritical pressure cooker’ hypothesis, and general confirmation of the cell phone hypothesis.

2. Child Care is Increasingly Regulated and Increasingly Expensive. Child care is required since you can’t leave your kid alone, and it is increasingly expensive. The idea that only 7% of your income should go to child care doesn’t make any mathematical sense even if you don’t require ground-floor rooms staffed by college graduates. If you do, what were you even hoping for? The solution seems to be that everyone pays way more than that, except sometimes the state covers it if you’re poor.

3. Let Your Children Play. Hand in hand with expensive child care is arresting any parent who lets their children out of their sight - or at least the risk that this will happen. There are no standards you can follow to be safe from this, at any age, for any activity, even at home. The mom arrested for letting a 14-year-old babysit is finally free, but her life is not going to magically be all right.

4. School Daze. Continued illustrations of what school is, and how we choose to run them. Examples of some very broken behaviors, even by our standards. A variety of common learning techniques don’t work.

5. A Case Against College. By trying to argue that college is worthwhile, Michael Story makes an excellent case that it is not. Those telling you to go to college increasingly embrace the signaling model of education. Claims that it ‘teaches you how to be productive’ better than alternatives don’t seem credible. The purported solution to disliking higher education is… more higher education.

6. Potential Alternative Methods. Is it possible to get much better school outcomes? To worry about your kids a lot less? To simply not offer enrichment?

7. Good Advice. A bunch of things parents worry about too much, you can stop. A lot of very central things make the list.

8. No More Fight Club. Some good news, fights are much less common now.

Why Are the Kids so Not Okay?

What additional thoughts have come in directly on the questions raised in The Kids are Not Okay?

Mostly I got strong agreement that our use of smart phones, scrolling and social media, and in general our relationship to computers and related technologies (note: not AI) are big problems for all of us, not merely children, and we are largely not handling these issues well. Those who do report handling them well report strict regimens of avoiding the toxic interaction methods, however they see those.

Some small additional notes that came in after press time.

Sarah Constantin points to college admissions pressures as an underappreciated source of teenage distress. Certainly does seem like ‘make kids feel like their entire future is on the line based on the shadowy decision of a rigged system, where the system actively punishes you if you don’t center your identity on calling it out for being rigged but demands you play anyway’ might screw a lot of kids up.

If college admissions isn’t on your radar it should be. The madness has gotten more intense, more obviously phony, and affects more teens (as more apply to college.)

For some reason it’s no longer discoursey but it didn’t actually go away.

“You have to be the very best to even be marginally okay in life”

“shadowy figures will judge your character and determine your future so you better pretend to be someone they’ll like”

“but also the game is rigged and if you were born into the wrong family you’re SOL”

Popular YA series these days like Scholomance are transparent allegories for this system. Y

ou will be eaten by monsters unless you get into a special wizard school and then into an upper-class “enclave”…in which case you have a 25% chance of not being eaten by monsters.

Also, as part of the process of trying to get into college, you now have to endorse a type of leftism that explicitly says “the game is rigged and winners don’t in any sense deserve to win”.

This is crazy-making in itself. Like, think about it.

Imagine a nice kid with any level of idealism.

“Ok, so if this is an oppressive system where some people get special privileges that they basically shouldn’t have…why do I even want to be an oppressor?”

Parents: “Because otherwise your life will suck even worse.”

Kid: “ok, that’s depressing, I guess the world sucks, but I get it.”

Kid: “wait, no, I don’t get it. Why are the same people who are telling me the system is phony and unfair ALSO trying to get me to join it?”

Kids understand the concept of a bad guy.

A bad guy who *wants you to affirm he’s a bad guy* but *wants you to act like you haven’t noticed somehow*???

That doesn’t even exist in fiction! It’s bizarre. Even if you don’t notice explicitly it’s got to be unsettling.

Sort of maps to: “Crucifying Jesus was the worst thing ever; the torture and murder of a perfectly good being.”

“So the Romans were bad guys, right?”

“Nooooo….”

“we ALL killed Jesus”

“ok I get that lots of people abdicated responsibility or failed to help or w/e but…shouldn’t we pay any attention to the guys who actually really did routinely torture political prisoners to death?”

“nope!”

Same kind of pattern of “you’re clearly describing evil and emphasizing how bad it is, and then expecting me to sorta…not react negatively at all to the particular people responsible?”

Derek Thompson at The Atlantic also focused on related questions, and adds ‘pressure cooker schools’ and academic competition to his list of suspects for increased teen anxiety, depression and unhappiness. Feeling obligated to do too much school work, feeling too much competitive pressure, increases reported mental health issues. As students academic achievements improve, they face worse competition because their aspirations also improve, and they get less happy. He notes in particular that:

• After an education reform in Germany led to more instructional time, a 2018 study found that increased school hours “significantly reduced adolescents’ self-rated mental health status.”

• A 2018 analysis of Taiwan “cram schools,” which often focus on helping students pass standardized tests, found that these programs reliably increase academic achievement at the cost of increasing rates of depression.

• A 2018 study by Gabriel Heller-Sahlgren at the London School of Economics found that schools around the world with higher test scores and more homework typically see the same trade-off: higher achievement and lower well-being, possibly because of greater “parental-achievement pressure,” which encourages students to feel more competitive.

• A 2022 study of seventh to 10th graders by Korean researchers found that students whose peers were randomly assigned private tutoring after school reported more depressive symptoms. One possible explanation is that teens without tutors became more worried about their relative academic success and spent more time on schoolwork to keep up.

Derek Thompson later wrote further at The Atlantic about the teenage mental health crisis, but it didn’t feel like this covered new ground. His podcasts on the issue are better, especially as he and his guests are then much more willing to express actual opinions, although I still wish he’d hedge even less. The more recent podcast emphasized even more the school pressure theory of the crisis, with a direct very large correlation between school achievement and teenage unhappiness across nations, and the rise of the ‘whole person’ college applications that rule out any possibility of having time to yourself without a threat to your future.

Freddie deBoar speculates on exactly how smartphones might make children anxious and depressed. Highly incomplete, some good additional angles.

Child Care is Increasingly Regulated and Increasingly Expensive

I think there’s an import link there, although it isn’t the whole story.

NPR asks about what looks like market failure in child care, with waiting lists everywhere, Arnold Kling proposes a potential two-part answer. First, the tax disadvantages of care reduce people’s ability to pay, although that shouldn’t prevent the market from clearing or cause waiting lists. What he thinks causes the waiting lists is that it is expensive to hold positions open.

Yes, you could raise prices enough that there were always vacancies, but you’d rather keep prices below that level to ensure full utilization, and also keep leverage over the parents - if you’re doing them a favor rather than the other way around, your job is a lot easier, and that keeps your workers happy. Kling proposes an online broker to find vacancies, same way we now have AirBnB, which could be useful but does not seem to me to be that promising in terms of solving the central issues. Sounds promising.

The key problem is that the tail risk of bad child care is super expensive, and quality is both important and hard to observe. When you try going on sitter.com or other such websites, there are some good options, but let me tell you, pickings be slim. If the online market were built around formal child care settings rather than individuals, hopefully that would help a lot here, including via normal reputation management?

One can also look at this from the standard regulatory perspective. Child care costs are driven up a lot, and supply pushed down, by various requirements like being on the ground floor or providers needing college degrees to care for little children, despite this making actual zero sense, if that was an actual real requirement all of us wouldn’t exist. This prevents supply from adjusting to meet demand - in a more competitive market you also wouldn’t have waiting lists.

Or it can be so much more stupid than that (Tweet includes 4 minute video).

You literally have to prove you are needed, are provided no hints what that would mean, and then they say no anyway 75% of the time. You see, this law ‘limits the burden on regulators.’

You know what would do an even better job of that? Yep.

Babysitting rates rose 9.7% in 2022 on top of the 11% hike in 2021. I can confirm. Strangely they think New York City rates dropped in 2022, to a level lower than Austin or Seattle, which I do not believe is accurate. The ‘why it matters’ section is concerned that ‘it’s also attracting teachers, nurses and other trained professionals into the career - which in turn drives up rates, because of their experience.’

Translated, higher pay rates are attracting skilled workers. To the extent that this higher quality increases willingness to pay, that drives up average pay more, but it doesn’t much drive up what you’d pay for the previous unexperienced labor. This is pure, good old reallocation of labor and supply to match demand.

The markets in question are highly inefficient, often very good providers work cheap while other not-as-good providers are expensive and still find work. Proper negotiating, networking and general strategicness matter quite a lot, on both sides.

The other concern is that the high cost of child care is ‘keeping some parents at home’ which, again, reflects supply and demand. If me going to work means someone else has to watch the kids, and the cost of that makes it not worth me working, then perhaps I should watch my own kids and someone else should work instead. Tax treatment can make this go the wrong way, where childcare payments are not tax deductible, so it isn’t foolproof. One wants a balancing act between the incentives, where the parent stays home if and only if they should, but in the long run I would likely lean towards more parents staying home on the margin, not getting more of them working.

The total cost here is a really big deal.

Per a 2022 Care.com survey, 51% of U.S. parents were spending over 20% of their income on child care — far more than the 7% that the federal government deems "affordable."

That 7% number is completely Obvious Nonsense. Watching children during working hours is a full time job. Yes, it is possible to use daycare as a force multiplier here to some extent, but if you have two incomes this implies a way to watch all your children for 14% of an individual’s income. If a typical worker to child ratio is something like 1:5 for infants and 1:9 for toddlers, and overhead and taxes exist, how exactly is that going to work even if the government isn’t imposing lots of expensive requirements that don’t do anything?

An intuition pump is that we want a capable person watching our children as a full-time job, and we want to be competent people working a different job, and somehow we want to each earn $7 working for each $1 we pay the people collectively watching all of our kids. What does that imply about ratios of caretakers to kids, and the ratio of parental pay to care worker pay? Do those numbers seem possible in general?

The solutions we are trying so far seem to be ‘impose lots of expensive requirements that do not do anything’ and then address the results with ‘massive government subsidizes to the poor so they can pay a lot more than 7% with Other People’s Money’ combined with ‘the non-poor pay a lot more than 7% with their own money.’

Either we Give Parents Money, or increasing numbers of people will decide they cannot afford children. The idea that we are simultaneously creating a set of taxes and regulations that make childcare and other family costs prohibitive, along with all the non-financial stress and requirements we add on top, and also ‘there is no way we could raise birth rates through government action?’ Absurd. Completely absurd.

Let Your Children Play

Almost everyone is failing to understand the real costs imposed by our absurd standards for supervision of children.

Here is another case study where parents were arrested for letting their children walk a few blocks on their own. In this case, the children were 6 and 8, and were walking to Dunkin Donuts in a quiet suburban neighborhood. Once again ‘sex offenders’ were used as the police justification. Once again, there was a child services investigation.

Woman who was arrested for letting 14-year-old babysit finally cleared of charges.

Neighbor writes in to the newspaper because they are concerned that a 13-year-old is left alone in their house on Saturdays, to ask if perhaps they should call child protective services about this? Yes, we are completely insane.

As opposed to when we used to be a proper country.

Jorbs: I think my parents were pretty happy that my primary school was only 30 minutes walk away because it meant they stopped having to drive me to school when I turned 5.

On the one hand, yes, this really does happen, and the threat of it has a huge chilling effect on behavior, including my own. Laws that clarified that such actions are fine, or at least were fine in the absence of particular reason for concern, would be a game changer in the lives of families.

Then again, these cases are news. If an individual case of something is news, that is an excellent sign that it is an unusual outcome. Is an arrest for letting your kids walk alone at age 8 similar to an airline crash? A bad thing with a very low base rate? Or perhaps is happens so rarely because no one dares break the rule? What matters is, conditional on letting kids walk around, what is the risk?

Or conditional on letting your child stay home alone, what is the risk? Even if that risk is minimal, how confident are you in that? Even if you are confident in that, are parents then willing to actually pull the trigger?

Once again, we need hard, clear, fixed rules for all this. There needs to be room for judgment calls and considering details, but also a safe harbor where if you follow reasonable rules you are fine barring showing of gross negligence - that parents are presumed to know what they are doing. There needs to be zero danger in letting a 14-year-old babysit your kids.

School Daze

Those offering services complain they will be hurt if competition is allowed.

From the comments on the last childhood post, on college grades:

So, I know for a fact that there are undergraduate programs at Harvard, Princeton, and Stanford in which the professors are implicitly (and sometimes explicitly based on a student’s donations to the school) forbidden from giving students realistic grades.

Also, the situation is only marginally better in other places. I taught at a major Canadian university for years and every year my faculty told me what the grade distribution and class average would be, and I would face implied consequences for not following this grading scheme. Also, I was very rarely allowed to fail someone, even when they absolutely deserved it.

After a number of years I finally just jumped ship and got a job in the private sector. Most university programs, as far as I can see, are just degree mills. Students pay inflated tuitions and receive their degrees. Any friction caused by the actual attempt to educate students is frowned upon and made to disappear.

Actually I could go on about this problem forever so should just stop here.

Also from the comments, on the question of certain high schools hiding national merit awards from students and parents, the national merit thing in question referred only to recognizing ‘pretty good’ performance rather than the big impactful top 1% awards:

Education Realist argues the National Merit thing is a big nothing. It was about some little known category that is meaningless, not the “real” semi-finalist status that actually matters.

https://educationrealist.wordpress.com/2022/12/28/whats-a-national-merit-scholar/

From the link, what they didn’t give out was the Commended Student award:

The essential category achievement in National Merit  ranking is “semifinalist”.

National Merit Semi-finalists are, roughly, those receiving the top 1% of PSAT scores. Designation isn’t an exact science, because the finalists are apportioned by state. Different states have tighter (California, Virginia) or looser (South Dakota) cut scores, but it’s basically the top 1%.

…

The other fact that jumps out is the 3%. Remember, semi-finalist is top 1%.  So Commended Student is way downstream from the only important category. Basically, a participation trophy.

… (last quotes here from the Part 2 post)

Commended ranking is not considered an award and does not render a student eligible for any scholarships they weren’t otherwise qualified for.

…

Semifinalist is useful to a [current senior] TJ student. Commended is not.

…

*Many colleges are now prohibiting test scores as part of admissions. TJ is also ending its test-based admissions so it will no longer be a given that the students are top 3-4%. The class of 2026 may value Commended status in a way that current TJ students do not.

This error did not extend to semifinalists and finalists, the awards that are super valuable and life changing. They also extended to other schools in Fairfax County.

That does not make the failure acceptable. It does mean the damage is limited, at least this time, assuming schools aren’t doing the ‘not being at least a Commended status student means you didn’t score in the top 3%’ move.

To me the key question here remains the motivation behind what happened. If this was intentional, or ‘accidentally on purpose’ by those who didn’t mind the error, it is rather awful. If it was regular incompetence, it’s not fine exactly, but it’s normal. Education Realist thinks it was regular incompetence. If this wasn’t at a school in the process of intentionally transforming from a special merit-based high school that competes for math championships to a regular school, I’d definitely agree.

Scott Alexander looks at various forms of schooling in his survey, warning that everything is of course confounded to hell and also ACX readers are not typical. Home schooling looks good before correcting for confounders, and contrary to standard theory the resulting children are not especially weird nor do they have social life issues.

(Unschooling had sufficiently low sample size Scott did not want to divide it into buckets, so despite it looking quite bad on the numbers mostly probably ignore it.)

When filtering out the religious the home schooling scores look less good.

So I selected the subgroup of respondents who are currently atheist or agnostic. This doesn’t rule out that some of the home-schooled children will have been religiously home-schooled, but they’re probably not benefiting from a religious sense of meaning and good church community now.

Here’s how this group rated their satisfaction with their schooling:

Public: 5.63
Home: 6.62
Religious: 5.51
Private: 6.50
Unschooling: 5.75

Here’s what life satisfaction looks like among this subgroup:

Public: 6.49
Home: 6.33
Religious: 6.53
Private: 6.56
Unschooling: 6.29

Now it’s lower! On to social satisfaction:

Public: 5.66
Home: 5.63
Religious: 5.85
Private: 5.96
Unschooling: 5.64

About the same for public and home; private and religious seem slightly higher. On to percent single:

Public: 40.4%
Home: 48.6%
Religious: 41.1%
Private: 42.0%
Unschooling: 50%

This looks pretty bad for home schooling. But remember, the average home schooled person is 29, compared to 33 for public school. In this subgroup, they’re even younger - the average atheist/agnostic home schoolee is about 27.8.

I worry this is somewhat unfair to home schooling, since losing one’s religion is not a great sign for life satisfaction or the successful passing on of values. Still, the figures here are noteworthy. These differences are small enough that this effect could explain things.

Also, of course, the main thing private school tells you is that the parents were probably at least upper middle class and likely rich.

Here’s where members of each social class said they went to school:

Poor: 81% public, 5% home, 8% religious, 4% private
Working: 78% public, 5% home, 11% religious, 5% private
Middle: 75% public, 3% home, 11% religious, 9% private
Upper-middle: 64% public, 2% home, 13% religious, 19% private
Rich: 42% public, 0% home, 16% religious, 39% private

Unschooling was too small a sample size for me to feel good including it.

Given this, despite students not being happy about their government schools, their test scores look remarkably fine here, as does life satisfaction. This updated me positively on public schools.

Answer broken down by schooling type:

Government school: 5.63
Home school: 7.04
Religious school: 5.87
Private school: 6.44
Unschooling: 6.00

As an additional measure of educational quality, I also asked respondents their SAT scores. For verbal:

Public: 743
Home: 756
Religious: 751
Private: 758
Unschooling: 758

And for math:

Public: 743
Home: 722
Religious: 727
Private: 757
Unschooling: 688

I wanted to check how people from various schooling types did in life - we’ll get to confounders in a moment. Here’s current self-rated life satisfaction, 1 is bad, 10 is good:

Public: 6.56
Home: 6.72
Religious: 6.65
Private: 6.62
Unschooling: 6.03

The catch here is that this is among ACX readers, which is a huge selection effect as seen by the SAT scores. I would like to see all of these numbers controlling for household wealth. Can’t take the survey too seriously due to the initial selection issues, of course. Still interesting, especially since ‘people who read this blog’ are not so far from the sample in question.

Study finds a lot of common learning techniques don’t work.

That sounds a lot like ‘if you want to learn X, practice X and test X.’

Ethan Mollick suggests using AI to create self-tests and help with expression, which makes a lot of sense. I am increasingly optimistic about the ability of AI to supercharge learning, when someone’s goal is to actually learn something.

Roman Helmet Guy has an alternative suggestion. Many people are saying this.

Time tested, taxpayer approved, enforced at barrel of a gun.

It seems part of that solution, at least in SF middle school, is to instruct teachers that when one child is assaulting a much smaller child, they are not to physically intervene, or inform parents of whether the perpetrator was even suspended?

a hearing impaired child was beaten viciously, unprovoked at my tween's SFUSD middle school, by a kid 5x his height and weight. nobody intervened to stop it, even though a staff member was present, because SFUSD instructs their teachers not to intervene. the hearing impaired child had to be taken to the emergency room. an email was sent out to parents days later saying a child was taken to the ER, with little or no info on the perpetrator. SFUSD protected the privacy of the perpetrator over letting parents know if their children were safe or if the perpetrator was suspended. SFUSD only does what it is minimally required because they weigh being sued over doing what's right and safe for 99% of all the other students. sound familiar?

It is essentially impossible to fully ‘protect the rights’ of students via a set of rules, because those rights will clash - if you can’t touch a student or reveal their status, then they go beat another student to a pulp, what do you do? Same as with adult rights.

New York City schools used to let teachers grade their own students’ Regents exams. For some odd reason, they stopped. Then for some odd reason, they decided recently to resume this practice. I can’t imagine why any of that happened.

The Ivy League perhaps is not so different. Here is a story claiming that cheating there is also rampant. When Columbia tells students to take a Calculus midterm from their dorm room, with no monitoring at all, and gives them 90 minutes of their choice - so students can even take the test at different times and talk in between - what exactly do they expect to happen? There is some point at which cheating becomes the norm, and even I can’t get angry about it.

A Case Against College

Michael Story, a very good predictor, has a Twitter thread that makes the case against college by attempting to make the case for college.

So far this is 100% a signaling model of education. Everyone burns four years and tons of money but do it or you’ll be left behind. If this is the central situation then we should ban colleges. Or, at least, ban discrimination on the basis of college.

Here is his case for there being any utility at all:

That is some giant weaksauce. You know what else makes you more productive and gives you better habits, only much more effectively? Actual work. No, really. I find the suggestion that college does this better than working, let alone the much better starting a business or doing useful things, absurd. ‘Luck surface area’ depends on what that means but if college is better at it then I’d say it means nothing we want to encourage, and says ‘luck’ is about social games.

Thus, try to improve this wasteful thing that accomplishes nothing? Or if you also find it too painful, use the fact that it also doesn’t take up too much time to supplement it, and power through until you’ve spent the years?

Even better, if you can’t stand university, the solution is more university.

This is the worst part, the idea that 18 year olds can’t be productive. If that is true, it is entirely because of college. Otherwise there is nothing stopping you from starting a business, or getting a job, or actually learning useful things.

Stanford hates fun but it loves money and price discrimination and 100%+ marginal tax rates.

The total undergraduate cost of attendance for full-tuition-paying families will be $82,406 in the 2023-24 academic year, with $61,731 for tuition and $19,922 for standard room and board. 

The new $100,000 income threshold for full financial assistance is comparable to Princeton University, where undergraduates whose families earn less than $100,000 annually will also receive full financial aid starting fall 2023.

Earlier this month, Yale announced that its total term bill will be $83,880 for the 2023-24 academic year. Yale’s threshold for full financial assistance will remain at $75,000.

‘Enough support to cover tuition’ is code for ‘we will take every penny you have, pray that we do not miscalculate and take pennies you do not have.’

I am fine with all this, actually. There will still be far, far more demand to go to Stanford than slots in Stanford, even if one excludes the unqualified. Why shouldn’t they charge families every penny they have (and likely many they don’t), and let the ones who don’t want to pay opt out?

Especially now that everybody knows Stanford hates fun, so we know the money is going to an investment in networking and signaling effects rather than a good time.

Don’t hate the player, hate the game and those who set the rules.

Potential Alternative Methods

Claim that a school in Austin is teaching kids academics with software, spending half the day on ‘life skills,’ and seeing 99th percentile outcomes. The first note is that admissions is selective, and looks for self-directed learners, so one must be extra skeptical. It still does seem like this should be possible - you can goal factor to learn academics in a way that does not waste most of your time, then use the rest of one’s time to learn the other things school is supposedly teaching.

There’s also this:

If you want to check out what they’re using and report back:

I am definitely curious.

Another group that learns nothing are students who already know what is being taught. A New Hampshire legislator proposed a bill to let gifted kids test out of high school, tells the story. No luck this year, or even next year. He still intends to try again. The opposition to such bills makes no actual sense, if you already know X being forced to sit in a room where X is taught is crazy. Except, what if X isn’t what is being taught? What if the cruelty and capriciousness are the point? To teach them that they are not so special, not to challenge power, that all men must serve?

DeSantis threatens to eliminate advanced placement (AP) classes. This seems like a highly stupid and destructive move on both political and educational grounds. AP classes are very good, as far as I can remember and tell. AP tests are mostly very good tests, again as far as I can remember and tell. They lower college costs, allow high school students to avoid being bored to tears and maybe learn something. All of this to pick a culture war fight over some little bit of the material.

Of course, if he did do it, it would be good for equity, which the WSJ reports is the reason many school districts are eliminating honor courses. Straight up Harrison Bergeron - the people are not equal, so we must harm those who are better off so we can fix that. Notice the implications for AI alignment.

One theory is that those who choose not to be parents often do so because they feel they would need to sacrifice too much, while those who choose to have kids sacrifice too little. So instead these non-parents, who would do a good job even without sacrificing so much, instead have no children and sacrifice nothing. Leaving all potential children, collectively, much worse off.

Good Advice

A thread of Good Advice for parents on what not to sweat, which is quite a lot. Here are some of the things I agree that you do not need to do:

1. Have everything be educational.

2. Constantly be teaching, including basic stuff they learn anyway.

3. Stop them from being bored sometimes.

4. Prevent social conflict.

5. Revolve your life and schedule around them.

6. Be their goofy buddy.

7. Buy infinite toys.

8. Keep them away from the real work.

No More Fight Club

One way in which things have changed a lot is the decline of physical fights.

This is not an unambiguously good thing, especially if this means that fights take other forms and happen more often because no one fears physical escalation. If fighting is unthinkable, beware the person who learns to think it. Or the models of those who do not understand that power ultimately still comes out of the barrel of a gun. This could also mean that when things do escalate, no one knows how to do so in a relatively safe manner that keeps guns and knives out of it and minimizes risks of permanent injuries. Sparring in martial arts classes seems to me to definitely be net good. The optimal number of physical fights, even real ones, is not zero. This is still is presumably mostly a good thing. Certainly if we scroll further to the left, we will see a lot of fights, far more than optimal.