Comment by camgunz

Thanks for the very thoughtful response! I admit I also get salty, particularly when it comes to education. It feels so obvious things should be a specific way, but of course that may just be my STEM person arrogance :P.

So, what I'm mostly confused about is why expulsion wouldn't work. We know some schools are better than others. We know students in "gifted" classes do better than others, and if your references are correct even a regular student in a "gifted" class would soak up the positive climate and turn out better than in a regular class. This seems to imply that expelling enough students should make the school better. For an extreme example, you could have everyone take a test, expel the lowest 50% of marks to a lower-tier school, and the remaining students would have better marks. This comparison is a little unfair, because expulsion is usually reserved for disruptive behaviour, not poor marks, but you could similarly have every teacher compile a list of misbehaving students. When I hear that expulsion wouldn't fix the problem, it must be because they are not expelling enough people!

I'm also a little leery of drawing the same conclusions as the news articles you linked. It seems likely that suspension/expulsion does always work, there's just a causation between lots of students misbehaving in a school and more students being expelled in the school. For example, the second news article says

> The findings underscore that suspending students does little to reduce future misbehavior for the disciplined students or their peers, nor did it result in improved academic achievement for peers or perceptions of positive school climate.

The linked findings come from this study:

https://www.air.org/sites/default/files/2021-08/NYC-Suspensi...

which has a few paragraphs on peer spillover effects from out-of-school suspension vs. in-school suspension. They do find a 1-2% decrease in the peers achieving ELA/math credit with out-of-school suspension (20-30% for the suspended), but there are also 20,000 incidents of out-of-school suspension with a median length ~two weeks [Table A.4]. Their data comes from the NYCDOE which has just under a million students, which means their peers also being suspended could account for half of the decrease! Then there's the correlation between negative school climate, more grievous offences, and out-of-school suspension (re: Table A.4), and it seems to be a clear-cut case of Simpson's paradox.

-----

Alright, time for the spicier part of this comment.

> I just wrote a whole screed in that Paul Graham wokeness thread about this exact thing. Educators are smart! They run studies on how best to educate! They're so easy to find and read!

I don't think the so-called educators are being smart. I think the average wokist is smarter than the average MAGAt (by a lot), but most systems fall into the Goodhart trap. People who optimise for looking good rather than being good often bubble to the top. This is why I think many woke arguments lean heavily on emotional appeals. The callous or ignorant MAGAts that only care about the gas price ironically end up with a more meritocratic system, because results matter.

I didn't partcularly like Graham's essay either, but I do sympathise with the anti-woke sentiment almost entirely because I believe this Goodharting has devastated the education system. For example, a common refrain I found in the comment section and your linked articles was,

> Expelling the student is not a good solution. Think about how this will effect his life! And what if he's going through abuse? Is it even his fault?

The MAGAt mentality is "I don't care, show me the results". They find current schools lacking, but don't particularly care about why they're lacking which is where school choice/vouchers come in. You don't need to fix things if you can just let the market find something better. This is a rather callous/ignorant take, and you can do much better by caring to find where the current system went wrong. I suspect it's because wokists forgot why we assign moral blame.

I think the purpose of "blame" in society is to figure out who to punish/rehabilitate to make society better. Note that even if there is a confounding factor it does not excuse the blame. I believe I've already mentioned this to you: you assign moral blame based on KL(bad action, person's policy). Why? If someone puts a gun to your head, and tells you to rob a store, you are unlikely to repeat the action. Your policy is really only, "rob stores when I have a gun to my head". On the other hand, if you were abused as a child and turned out a kleptomaniac, you are extremely likely to repeat the offence.

Now, rehabilitation has to actually work. If someone is starving, it doesn't matter how many beatings you give them, they will still steal food. Positive rehabilitation is often better for society, because you don't need to spend a bunch of money on the justice system, and the rehabilitated criminal can hold a job and pay taxes. Punitive rehabilitation works by decreasing the cost of future crimes from similarly-minded people. Note that I'm being really careful to talk about what is good for society, not the criminal. After all, every individual except the criminal (and friends/family) gains more by asking for the good of society, not the individual.

This ties into wokism and education as so: the wokist gives the emotional appeal,

> Expelling the student is not a good solution. Think about how this will effect his life! And what if he's going through abuse? Is it even his fault?

and the proper response is,

> good for whom?

As I mentioned at the start of this comment, it is good for the top 50% of students to expel the bottom 50% to an alternative school. Should we? In reality, we have to work under money and (as you pointed out) pitchfork & torch constraints. My issue with emotional appeals is they bring out the pitchforks, for potentially no good reason.

For example, I went to middle school in a rather conservative city, but even there the gifted program was eliminated in the name of equity. High school graduation standards have dropped, again in the name of equity. And California briefly proposed not allowing 8th graders to take algebra (for equity's sake) until they received massive backlash.

I care much more about what is actually good for society than what looks good. I really don't see how it's good to be holding back our brightest students to the bottom quintile's pace, or allow disruptions from known troublemakers.

At this point I'm rather tired; I might continue writing this tomorrow, but I probably won't. I'll just end with what I wish the school system looked like:

1. A national placement exam for each grade (including Kindergarteners). Students get placed into schools and classrooms entirely from their rankings (within the local system). The top scorers are offered room and board at nationally-run schools.

2. Disruptive students get kicked to penitentiaries. I read elsewhere in this thread of a city with three tiers of schools: one for regular students, one for first-time expulsions, and a last for the chronically expelled. This is what I'm imagining.

3. The same people that write olympiad problems and run the summer camps are hired to create a new curriculum. Quite frankly, Common Core is a failure; you see a decline in AMC 10/12 scores and participants about eight years after it was introduced, i.e. just enough time for the students who learned from Common Core to be taking the exams.

4. Everyone is fired, and as many people as money there is are hired at $300–500k/yr (in total compensation) to teach. At $15k/yr per student (what the US currently spends), and 30 students to a class, this should be just doable.