This all seems clearly down stream of the problem that the justice system does not have the resources to take even 1% of criminals to trial. Criminals really really do not want to spend time in prison, and are unwilling to accept plea deals that involve prison time. Prosecutors, knowing they cannot go to trial, bend over backwards to arrange plea deals that make the criminals life miserable in other ways but do not actually get them off the street. Police officers, seeing the same person arrested for the same thing for the 10th time this year, stop bothering to even make arrests for minor offenses. The definition of "minor" continues to creep upward.
This will continue until there are much fewer crimes or the state develops the capacity to actually force punishments on criminals who do not agree to them.
Absolutely. I would also be interested to see the cost required to try a single case over time. I’d imagine it’s exploded. McKinley’s assassin was executed 45 days after the assassination. Literally caught, tried, convicted, and executed in a month and a half. It seems to me that we’ve created a system that is just unworkable. The expectations around procedural fairness and thoroughness of a defense have risen so much, and I’m not sure it’s been worth it - not to mention that the reinterpretation of the “right to legal defense” as a “right to state funded legal defense” has caused more of this burden to fall on the state rather than defendants.
re: "This all seems clearly down stream of the problem that the justice system does not have the resources to take even 1% of criminals to trial. "
Well.... if a small # of people are committing most of the crimes, (as the data in Scott's post suggests) then actually it seems like that is the root problem. If now we can't afford to take even 1% of criminals (which by definition would be largely career criminals getting constantly re-arrested) to trial, then it seems like the solution is to take those people (we can easily ID them, they've been arrested 17 times) to trial and give them long sentences so you reduce the future pool of "people we might have to take to trial" and can therefore focus on the important ones.
Something to keep in mind that makes a lot of these studies hard to detangle or apply common sense to is that different studies define recidivism differently, and it may also depart from an individual's common sense definition. It's similar to "assault weapons" in that way.
Typically if someone is convicted for robbery and is released on community supervision (probation, or parole after a prison sentence), they have additional rules to follow. Violation of those additional rules can lead to rearrest, reconviction, etc. This would generally be considered part of the recidivism metric. It does indicate a certain level of not-rule-following that is important to consider. However, it is not the same thing as being arrested for another robbery.
It would be nice to have a study that focused on reoffence, and even nicer to filter out various "victimless" crimes. I know dependency can create a vicious cycle but it would be interesting to see the difference in reoffence between robbery -> robbery / drugs -> drugs / drugs -> robbery / robbery -> drugs.
Did he say that? I thought Scott's stance was that Ivermectin worked as a dewormer and so it improved outcomes mostly in places with high strongyloides ( or similarly susceptible parasite) burdens.
He belatedly and with great trepidation did a follow-up to try and address Marinos Luther's 95 Ivermectin Theses, and...I'm glad it was attempted at all, plus it did charitably concede some errors and genuine valid counterpoints. But the heroic level of effort necessary to properly disentangle that epistemic web was beyond what Scott wanted to offer, so he kinda half-assed it overall. Which I frankly don't blame him for, especially given the unending barrage of drive-by commenters showing up on random posts/Open Threads to pimp for Alexandros and try and bait Scott into """debating""". Which had a distinctly religious flavour, lots of dire implications that This Is A Black Mark On The Entire Rationalist Movement, A Turning Point Marking The End Of SSC, Suppression Of Truth, etc, etc...glad that mess is in the past now, one way or another.
Even Markos seems to have finally given up, after a year or two of attempting to gain Scott's attention. But some of his acolytes don't realize the stand-down order was given and keep on embarrassing themselves.
Wanting to re-litigate the one moment of relevancy you ever had is something to talk about quietly with your therapist, not to post openly on every single comment thread you can find.
"Our current prison system is really bad in that many aspects cause more crime after release rather than less, and the low hanging fruit is fixing this so that it isn’t true."
The solution seems pretty obvious. A small number of people commit almost all crime. Identify these people (3 strikes?) and lock them up, or make them wear a monitor bracelet or go to some rural work camp (I don’t care where they go, just as long as it isn’t near the rest of us). If they commit a fourth strike after getting out, lock them up until they qualify for a retirement discount.
If this takes more prisons, more cops, better courts or whatever, then let’s spend money on these as needed. That is something government should spend on.
One paradox is that crime is a power law distribution. So yes, it's true that 300 people account for 70% of the shoplifting in NYC; but if you arrested those 300 and sent them to a labor camp, it would _still_ be true that 300 people account for 70% of the (remaining) shoplifting in NYC.
While this pushes the limits of my ability to mathematically model outcomes, it's my understanding that crime may not follow a *self similar* power law, meaning that removal of, say, a very prominent gang could very well alter the distribution of crime and not necessarily lead to a newer, smaller gang emerging or existing. Disrupting gangs or similar activities can change the fundamental average nature of criminal perpetrators, altering the distribution.
Let's look at a company productivity where sqrt(EmployeeNumber) creates 50% of the productivity. 100 employees and B amount of work, 10 employees does B/2 of the work.
Remove those top 10, now the remaining 90 employees do B/2 total work, and Sqrt(90), or 9 emplyees and two severed legs does B/4 and the remaining 80 employees with corpse legless trunk does the other B/4 amount of work.
Take the top produces away, only 1/2 the production is done.
Now if those 300 shoplifters were all that could fit in store and 10,000 more possible shop lifters would shoplift but not able to fit inside, then remove 300 would only free space for 298 and Fat Billy in to shoplift. So, yes, then your power-law example would be correct, .. except if those remaing 10,000 shoplifters were migits with tiny hands that could not carry as much. Poor migits, don't throw them!
Some people are so dumb that prison doesn't deter them - and the answer to this is to not send these people to prison as much? Completely backward.
Really, it means some people cannot be trusted to not be removed from society. Like, do people not understand that *You can't do crimes when you're in prison*?
There seems to be a one to one overlap between YIMBYs and people who GROSSLY discount the social cost of crime. Which is really, really hilarious considering that crime ridden inner cities is an enormous part of why there is so much sprawl in America and why people don't want to live in the city as much as other parts of the world. Of course, YIMBYs are either too dumb or too ideologically blinded to put two and two together and would rather just call white people "racist" - apparently we have a moral obligation to put up with massive amounts of violent and property crime instead of trying to protect ourselves and our families from this at enormous personal cost.
Of course, they also assert that government programs will fix everything in spite of a completely lack of evidence for this beyond assuming every correlation that matches their ideology is 100% causation. Because hey, maybe if we give books to black families, black kids will start doing better in school, right?
In fairness, it's worth asking what methods best deter crime. Property crime seems to cycle with government food assistance (Meaning crime is higher when assistance has been less recent), suggesting that property crime is need driven. An average cost of a year in prison is about $38,000 to the state, with some states being much higher. This doesn't include the cost of law enforcement or prosecution, much less crime victimization.
On average, SNAP participants will receive an estimated $187 per month
So if giving SNAP to 17 people would prevent one person from being in prison for a year then it could easily be better to expand assistance rather than to increase prosecution for property crimes.
Of course, this is a very simplistic model that makes some taboo tradeoffs and doesn't account for feedback loops. Even so, it's worth at least asking: If we're going to spend public funds to reduce property crime, what's the best way to do it? While violent crime seems to correlate with socioeconomic status, short term economic assistance doesn't seem to deter violent crime, that I can see. (Improved nutrition does reduce violent crimes, especially when forced in controlled environments like prisons, but there seems to be a disconnect between increasing supplemental food assistance and improving nutrition.)
Matthew Yglesias, Official YIMBY Spokesperson + pariah on the left due to tough-on-crime, anti-woke, pro-private market views, would like a word. But don't let evidence stop you from righteously dunking on an outgroup, yeah? (Nevermind that Zvi is also YIMBY-pilled?)
Personally, I'm half-hoping for a "natural experiment" where...vigilante justice...is served to the minority responsible for All The Crime, and then suddenly all the whining stops because everyone finally realizes Crime Is Bad, Acktually. While I Do Not Condone This, if the government won't protect the people, then the people can't be faulted for protecting themselves...
Thanks for this. Just a note that when I studied theories of punishment in law school, the prof separated "deterrence" (penalties that prevent otherwise-might-be criminals from committing crime) from "separation" (penalties that remove convicted/repeat criminals from society, thus preventing them from committing more crime). I think it's a useful distinction instead of referring to both under the blanket name "deterrence."
On funding the police more - you also have to couple that with rules or laws that they can't just use it on traffic stops. I did a post analyzing this, and ~80% of police time goes to traffic stops and overhead, and this is while clearance rates for murder are ~50% and rape and assault <30% and property crime <10%.
Even if all we did was mandate "you have to spend 90% of your current "traffic ticket" time on actually solving crime," we'd roughly triple the amount of police hours put into actually solving crime.
The current allocation of police hours is absolutely bonkers relative to our clearance rates of serious crimes, and we should not be giving them more money while they're going to keep allocating time in bonkers ways.
> Don't traffic stops consistently uncover people with outstanding warrants, firearms violations, etc.?
If nearly all discretionary police time is going to traffic stops and closure rates are still <50% on everything important, just how many uncovered warrants and crimes can it be uncovering?
Clearly not enough, and spending time on actual police work might be in order.
I think there's something to what you say, but seems worth noting that "conduct traffic stops" and "investigate murders/rapes/assaults/thefts" are different skillsets, and presumably there are lots of cops who can do the former but not the latter.
Sure, but if we know they're wasting the vast majority of police hours while not solving any real crimes, it's a clear sign they're bad at allocating their resources, and we shouldn't give them any more without fixing that.
Obviously, do something like hire one detective instead of 2 or 3 traffic cops if it's a money issue, and we'll all be better off.
The world doesn't need 2-3 more traffic cops, that niche is filled and pointless anyways, but it does need 1 more detective who will actually solve crimes, and this writ large across all the cities will be a better end state for everyone.
I certainly would prefer a better allocation mix of fungible police man-hours and money (granting that they are, indeed, fully fungible). However, given how local traffic law enforcement cratered during covid and still hasn't reverted to the mean, I've gotta quibble with that pointless niche being filled. It's genuinely a worse equilibrium for me to live along a road that's now used for drag racing and other reckless driving on the regular! Some tickets are bullshit revenue-padding, sure, but it's not like moving violations aren't also crimes in their own right. (Perhaps a compromise would be diverting traffic cop money to red light cameras and other such surveillance? Probably more efficient anyway...)
"If we value a life at $12 million when calculating health care interventions, you’re telling me marginal murders only have a social cost of $9.4 million? That’s crazy, murder is considered much worse than other deaths and tends to happen to the young."
1. Not sure where you're getting these figures, but the $12 million figure might come from the value of a statistical life, which peaks around 46 years old, because it assigns the highest value to middle-aged adults based on economic productivity and willingness to pay for risk reduction. If you're using QALYs, ICER uses $100,000–$150,000 per QALY, so that could go as high as 10 million for a young victim? 9 million feels right.
It seems to me that the most constructive answer to "should we adjust prison sentences to be shorter / longer" is "mu" – don't focus attention on this question. As you say, the low hanging fruit clearly lies elsewhere (and there is likely a lot of it). We should focus on having a prison system that doesn't cause more crime after release, looking for ways for punishment to be more certain and swift, etc.
I dunno, if you mean "easy" in the sense of "politically palatable" then I agree, it's quite a quagmire.
If you mean "easy" as in, "one can write a policy prescription that would work", then this seems pretty easy? Like, start with:
* Significantly increase police funding
* Possibly also increase court funding, if needed to keep up with police cases.
* In order to fund the above, either cut other parts of the justice system (e.g., prisons), or raise taxes, or both
* Get rid of police unions
* Bring back corporeal punishment
A harder thing to do, which doesn't seem that hard considering the currently shoddy state of things, is to improve the administration of police services. I'm thinking of things like,
* Identify the bottom quintile or so of police officers, and fire them.
* Figure out the right balance of number of officers, officer pay, equipment, training, etc.
One other thing based on the comments at Astral Codex Ten is that it's probably worth figuring out probation; the current system doesn't seem to work at all but I could imagine one that could work well at low cost.
1. for all but significant crimes then - Each local community would have a Judgment-Committee of 5 to 9 that are fathers or grandfathers that have raised children well that displays good prudent Judgment, and they would rule on local violations of good social behavior for adults 15-years and older. Such violations as defined to be against good social order, such as; lies, stealing, disrespecting social-standards, etc.
2. In today's society False-Witnessing, Lying, harming the reputation of others can not be corrected, so such abusers will never benefit in Loving Correction – a punishment to make clear they behaved badly and motivation to not repeat the abuse. Using Public Pain&Shame will pay the wages of their sin.
Punishment involves public pain and shame. Once a week, those to be punished would be taken to a public place like a park or public building court-yard and restrained to an anchor, and a collar attached to their neck, and for the directed period defined by the Judgment-Committee, the length of time that the collar delivers the taser sentence.
Those people that do not want to be punished need only to be respectful to those in their community and if they make a mistake they personally engage those they harmed and give satisfaction for the damage, if possible. If not, then the Judgment-Committee is used.
Those of age 14-years old would be required to view at least 3 of these punishments before turning 15-years old and be a risk of the same punishments for those community social crimes.
and decided to use it a template to replace prisons. What do you think?
Modern Rehabilitation Camps: A Vision for Humane and Constructive Incarceration
Introduction
The modern prison system has long struggled to balance punishment with rehabilitation. In a society increasingly recognizing the value of human dignity and social integration, a new model of incarceration must emerge. This model, which we shall call "Rehabilitation Camps," draws inspiration from principles of order, discipline, and the promotion of virtue. Its purpose is not merely to confine but to reform and elevate the men within its walls, preparing them to rejoin society as honorable and productive citizens.
Principles of the Rehabilitation Camp
The core principle of the Rehabilitation Camp is to instill a deep sense of discipline, purpose, and moral virtue in its inmates. By emphasizing skill-building, social integration, and humane treatment, these camps aim to transform individuals who have fallen into criminality into dignified members of society. This system is not designed to humiliate or degrade but to uplift and inspire.
Moral and Social Education: Every inmate will be provided with a structured education in virtues such as honesty, self-discipline, accountability, and charity. Daily routines will involve lectures, readings, and discussions on moral philosophy, social order, and the responsibilities of citizenship.
Work and Trade Training: Inmates will be assigned meaningful and productive work according to their physical and intellectual capacities. This work will range from agricultural labor to skilled craftsmanship, preparing men for productive employment upon release. Employment training shall include apprenticeships, technical skills, and creative trades.
Physical Discipline and Health: Physical health will be seen as an essential aspect of moral clarity and personal strength. Inmates will engage in daily physical training to promote self-discipline and bodily health. Balanced, nutritious meals will be provided, and medical care will be readily available.
Guard as Mentor and Model: One of the most critical features of the Rehabilitation Camp is the role of the guards. The guards are not merely enforcers of rules but are to act as mentors and exemplars of moral virtue. Guards will be selected from among the finest citizens — men who embody self-restraint, honor, and compassion. Their conduct must reflect the highest standard of virtue so that they inspire respect and emulation from the inmates.
Gradual Reintegration and Parole: The ultimate goal of the camp is not indefinite detention but gradual reintegration. Inmates who demonstrate genuine reform will be granted increasing privileges and responsibilities, eventually transitioning to supervised parole and employment placements. The emphasis is always on reform and uplift, not punishment for its own sake.
Management of Violent and Impulse-Control Prisoners
A key challenge of any correctional facility is managing individuals prone to violence or severe impulsive behavior. The Rehabilitation Camp will approach this issue with particular care, ensuring the dignity and safety of all inmates while guiding the violent man toward self-mastery.
Controlled, Isolated Environments: Individuals with serious impulse-control problems will initially be housed in separate, highly monitored sections of the camp. These facilities will be designed not for punishment but for intensive correction. The goal is to protect these individuals from themselves and protect others from their aggression, allowing space for gradual reform.
Intensive Counseling and Discipline: Specialized counselors and chaplains will work closely with these inmates, using a combination of religious instruction, cognitive-behavioral therapy, and moral reasoning to assist the inmate in gaining self-mastery. Guards will be particularly chosen for their patience and virtue when working in these wings.
Physical Labor and Discipline: Labor has a transformative effect on unruly men. Those prone to violence will be given the most physically demanding but meaningful tasks, such as agricultural work, stone cutting, or timber work. The structure of hard physical labor coupled with moral instruction has proven historically effective in fostering discipline.
Strict Accountability with Dignity: Violent outbursts or breaches of discipline will be met with swift but humane consequences. Isolation or suspension of privileges will reinforce the importance of self-control. However, physical mistreatment or dehumanization will never be permitted. Dignity is paramount.
Pairing and Team Integration: Once a violent prisoner demonstrates a measurable level of self-control and discipline, he will be paired with another prisoner of similar progress to form a working team. These teams will engage in collaborative physical labor requiring cooperation, communication, and mutual reliance. This method serves to increase social aptitude and reduce isolationist tendencies.
Expedition and Autonomous Projects: As teams of two begin working efficiently together, they will be further integrated into larger teams of four. Once these four-man units demonstrate reliability, they will be assigned independent projects outside the main camp. One such project will involve leaving the camp for two to three days to construct their own sub-camp using provided tools and materials. The men will design and build their own cabin, utilizing creativity and teamwork. While largely free from guard oversight during these tasks, the men must adhere to clear task completion deadlines, dedicating three-quarters of their daily work hours to construction, farming, or resource gathering. This autonomous period fosters self-reliance and continued social skill development, preparing men for eventual reintegration into society.
Gradual Trust and Supervision Reduction: Successful completion of sub-camp projects will result in increased trust and privileges. Ultimately, men who demonstrate consistent reliability, moral reform, and teamwork aptitude will be eligible for early parole, transfer to open-labor programs, and eventual reintegration into society.
Case Study: Hypothetical Criminal in Two Systems
To demonstrate the potential benefits and drawbacks of Rehabilitation Camps compared to standard prisons, consider the case of John, a 32-year-old man convicted of armed bank robbery. John, a divorced father of a four-year-old son, faces a maximum sentence of four years.
Standard Prison Experience:
Social Isolation: In a standard prison, John would experience severe social disconnection. The limited family visitation and lack of community engagement would likely distance him further from his son.
Minimal Skill Acquisition: With limited access to meaningful labor or skill-building, John would likely leave prison without marketable skills, reducing his employability.
Psychological Deterioration: The punitive environment, lack of positive role models, and inherent exposure to criminal subculture could worsen John's behavior or mental state.
Limited Rehabilitation: Upon release, John would face immense difficulty reintegrating into society, struggling with stigma, lack of skills, and damaged familial bonds.
Rehabilitation Camp Experience:
Social Reintegration: John would work in teams, engage with mentors, and gradually restore his social aptitude, maintaining a sense of community.
Skill and Trade Development: Through labor assignments, John would learn practical skills in construction, farming, or other trades, improving his employment prospects post-release.
Positive Role Models: Guards and staff act as mentors, demonstrating discipline and virtue, which would inspire John to reform his own behavior.
Family Reconnection: Gradual reintegration processes, increased communication privileges, and visitation rights would help John maintain a relationship with his son.
Higher Rehabilitation Success: The structured, dignified environment of the camp increases the likelihood of John becoming a productive citizen post-release.
In conclusion, while standard prisons rely heavily on punishment and isolation, Rehabilitation Camps emphasize moral restoration, social reintegration, and skill acquisition. This model offers a clear path to societal contribution, reducing recidivism and restoring the dignity of incarcerated men. John’s case demonstrates how an approach grounded in virtue and productivity can transform not only individuals but also the communities they return to.
In this way, the modern prison system may evolve from a mere instrument of containment to a powerful engine of social and moral uplift, truly embodying justice and mercy in equal measure.
There are a few factors I'm surprised you didn't mention - though I'm sure this was already a lot of work.
Firstly, how much correlation is there between things that are (illegal + measured), and things that are (actually happening + harmful)? Eg, 327 professional shoplifters getting the catch-and-release treatment 20 times a year is obviously bad; but wage theft dwarfs all "conventional" forms of theft combined and is also punished by a slap on the wrist in the few cases it is even reported. A shoplifter may know they risk being arrested, inconvenienced for a day, and released; a manager committing wage theft knows perfectly well they will never be arrested and at worst the company they work for might be fined. Might it be more cost-effective to work on reducing wage theft?
Secondly and similarly, civil forfeiture laws have resulted in tens of billions of dollars being seized by police departments without due process, and the majority kept for their own use. Many of these seizures are never contested, because paying a lawyer to prove you *didn't* commit a crime costs more than the $2000 that was stolen. Some of this money is presumably taken from criminals, but to my knowledge we have no good way of knowing how much. This is a theft that is by definition committed only by police departments and could be much more readily stopped than chasing down individual criminals, trying, and imprisoning them.
I think this gets to the "victimless crime" distinction. Siphoning money is a widespread aggregate harm, and if we're just talking cents and dollars, it's [magnitude] more than the value of actual physical stolen goods. On the other hand, no one ever got shot during a botched wage theft, and there is indeed a qualitative difference when the stick-up kid wears white gloves, so to speak. Even just in terms of societal norms...Everybody Knows books get cooked, people fudge punches, veteran employees are mysteriously terminated right before qualifying for pensions. None of that erodes my faith in collective humanity, just individual managers and companies cashing out their integrity cheaply. But that's a different class of values than the idea that it's okay to just take stuff without paying, than believing that the average person would totally do a thieving if they could get away with it. Shitty for the employees too, I miss the days where I assumed the best of my customers by default and didn't constantly catch myself statistically discriminating. Power flows from the barrel of a gun, and all that...we don't actually want to go back to a world where what's yours is only so because you keep others from stealing it. So I think some amount of extra scrutiny/resource allocation is merited on the grounds of norm defense, above and beyond just the dollar value.
(Totally amenable to slapping wrists harder, of course! Chewing and walking gum! I just think it wouldn't be ideal at a direct 1:1 rate to trade e.g. shoplifting prosecution for wage theft prosecution. One must always solve for the equilibrium, as well...it seems pretty clear by now that recent spikes in shoplifting and other petty crimes were largely due to increasing awareness that there *isn't*, actually, much risk of getting arrested, let alone meaningfully punished. Paper tiger enforcement doesn't work so well once the cat's out of the bag. Considering how fast that spike was, I'm really cautious of ceding more such ground...it's so, so hard to recapture. Whereas wage theft is an endurance race, and the thought of seriously addressing it under the current administration anyway is...aspirational.)
Civil asset forfeiture is mostly bullshit though, yeah. Even to the extent it hurts actual criminals, it's like KYC-AML, where the squeeze far outweighs the juice...haven't found a single compelling defense of the practice. Yet it persists anyway, so...qui bono?
Ah, a glimpse into the AGI Normal timeline, where Zvi blogs about "mundane" topics...
Timely post to read after a 5/10 harrowing late-night train ride home with a crazy high guy (really high, or high as well as crazy? guess it doesn't matter)...and the whole time I'm thinking, ugh, I just want to decompress on my commute home and not worry about shit like this, can we not, why is it so hard to deal with even the Obvious Ones. That's the social costs angle: even though nothing serious actually happened, there's clearly a lot of counterfactual value being set on fire, for everyone raising guards on the train + waiting on platform. Big Shut Up and Multiply energy. I'd pay a lot/have to be paid a lot to (not) be assaulted. And even more to go back to societal norms of mostly not having to worry about any such interlocution at all, of course...which wasn't even that long ago! ~Six years is within living memory of just about everyone!
Your methodology note is interesting...does it vary by question, or is it a more general shift? Criminology is well-known to be riddled with shitty epistemics and confounds (hell, even getting reliable data on actual non-murder crimes is hard), so I can understand being extra skeptical of the classic More Than You Wanted To Know treatment. Academic GIGO is still garbage, there are alternative ways to accurately map the territory. But some of your other notable "takes" posts like Contra Deceptive Car Seats, Jonestown Act Massacre, Never Sports Bet Against A Sicilian, etc. do rely on studying the studies? I guess the treatment is different even when working from the same epistemic base though...still enjoy Scott's analyses, but he seems (increasingly?) too nervous about rolling to disbelieve, deriving from first principles, etc. Admirable approach, falls short when the Official Data just isn't there/too contradictory. It's hard to live one's life on responsibly uncertain hedge bets, sadly, especially for immediately-relevant issues like crime and punishment that are actively being contested.
Have your sons join the HS westling team, and your daughters take up the batton and gymnastics. The batton handing skills will unknowing turn her into a deadly weapon handler .. I know my older sister learned the batton and darn hear knocked me unconsious not realizing how serious a blow she game me.
And if your sons get confortable in public battles with strange young men his size and if he wons most often, your sons will make soild eye contanct with the psychos on subway or the gang of three and they will know they should target some one that woundn't maybe F-them-up and then when they down stomp on they necks so they never ever target anyone else - back to Dust with You!
But tell your sons when they go to keg parties with pot to not get into any fights in parken lot and thow him around like rag-doll bounding him off the cars - because he might not hear Holy Mary's warning that he is about to kill a man, and that is Forever.
This all seems clearly down stream of the problem that the justice system does not have the resources to take even 1% of criminals to trial. Criminals really really do not want to spend time in prison, and are unwilling to accept plea deals that involve prison time. Prosecutors, knowing they cannot go to trial, bend over backwards to arrange plea deals that make the criminals life miserable in other ways but do not actually get them off the street. Police officers, seeing the same person arrested for the same thing for the 10th time this year, stop bothering to even make arrests for minor offenses. The definition of "minor" continues to creep upward.
This will continue until there are much fewer crimes or the state develops the capacity to actually force punishments on criminals who do not agree to them.
Absolutely. I would also be interested to see the cost required to try a single case over time. I’d imagine it’s exploded. McKinley’s assassin was executed 45 days after the assassination. Literally caught, tried, convicted, and executed in a month and a half. It seems to me that we’ve created a system that is just unworkable. The expectations around procedural fairness and thoroughness of a defense have risen so much, and I’m not sure it’s been worth it - not to mention that the reinterpretation of the “right to legal defense” as a “right to state funded legal defense” has caused more of this burden to fall on the state rather than defendants.
re: "This all seems clearly down stream of the problem that the justice system does not have the resources to take even 1% of criminals to trial. "
Well.... if a small # of people are committing most of the crimes, (as the data in Scott's post suggests) then actually it seems like that is the root problem. If now we can't afford to take even 1% of criminals (which by definition would be largely career criminals getting constantly re-arrested) to trial, then it seems like the solution is to take those people (we can easily ID them, they've been arrested 17 times) to trial and give them long sentences so you reduce the future pool of "people we might have to take to trial" and can therefore focus on the important ones.
Something to keep in mind that makes a lot of these studies hard to detangle or apply common sense to is that different studies define recidivism differently, and it may also depart from an individual's common sense definition. It's similar to "assault weapons" in that way.
Typically if someone is convicted for robbery and is released on community supervision (probation, or parole after a prison sentence), they have additional rules to follow. Violation of those additional rules can lead to rearrest, reconviction, etc. This would generally be considered part of the recidivism metric. It does indicate a certain level of not-rule-following that is important to consider. However, it is not the same thing as being arrested for another robbery.
It would be nice to have a study that focused on reoffence, and even nicer to filter out various "victimless" crimes. I know dependency can create a vicious cycle but it would be interesting to see the difference in reoffence between robbery -> robbery / drugs -> drugs / drugs -> robbery / robbery -> drugs.
Is this the same Scott Alexander who wrote that lovely essay about Ivermectin not working only to be eviscerated by substacker Alexandros Marinos
Did he say that? I thought Scott's stance was that Ivermectin worked as a dewormer and so it improved outcomes mostly in places with high strongyloides ( or similarly susceptible parasite) burdens.
He belatedly and with great trepidation did a follow-up to try and address Marinos Luther's 95 Ivermectin Theses, and...I'm glad it was attempted at all, plus it did charitably concede some errors and genuine valid counterpoints. But the heroic level of effort necessary to properly disentangle that epistemic web was beyond what Scott wanted to offer, so he kinda half-assed it overall. Which I frankly don't blame him for, especially given the unending barrage of drive-by commenters showing up on random posts/Open Threads to pimp for Alexandros and try and bait Scott into """debating""". Which had a distinctly religious flavour, lots of dire implications that This Is A Black Mark On The Entire Rationalist Movement, A Turning Point Marking The End Of SSC, Suppression Of Truth, etc, etc...glad that mess is in the past now, one way or another.
It's depressing: https://astralcodexten.substack.com/p/response-to-alexandros-contra-me
Even Markos seems to have finally given up, after a year or two of attempting to gain Scott's attention. But some of his acolytes don't realize the stand-down order was given and keep on embarrassing themselves.
Wanting to re-litigate the one moment of relevancy you ever had is something to talk about quietly with your therapist, not to post openly on every single comment thread you can find.
What's going on with the video linked to "As a marker, before I began reading the post, I put down here: Yes"?
Lol
"Our current prison system is really bad in that many aspects cause more crime after release rather than less, and the low hanging fruit is fixing this so that it isn’t true."
"Hanging" seems to be the key word there.
The solution seems pretty obvious. A small number of people commit almost all crime. Identify these people (3 strikes?) and lock them up, or make them wear a monitor bracelet or go to some rural work camp (I don’t care where they go, just as long as it isn’t near the rest of us). If they commit a fourth strike after getting out, lock them up until they qualify for a retirement discount.
If this takes more prisons, more cops, better courts or whatever, then let’s spend money on these as needed. That is something government should spend on.
One paradox is that crime is a power law distribution. So yes, it's true that 300 people account for 70% of the shoplifting in NYC; but if you arrested those 300 and sent them to a labor camp, it would _still_ be true that 300 people account for 70% of the (remaining) shoplifting in NYC.
While this pushes the limits of my ability to mathematically model outcomes, it's my understanding that crime may not follow a *self similar* power law, meaning that removal of, say, a very prominent gang could very well alter the distribution of crime and not necessarily lead to a newer, smaller gang emerging or existing. Disrupting gangs or similar activities can change the fundamental average nature of criminal perpetrators, altering the distribution.
Noooo, this doesn't seem correct.
Let's look at a company productivity where sqrt(EmployeeNumber) creates 50% of the productivity. 100 employees and B amount of work, 10 employees does B/2 of the work.
Remove those top 10, now the remaining 90 employees do B/2 total work, and Sqrt(90), or 9 emplyees and two severed legs does B/4 and the remaining 80 employees with corpse legless trunk does the other B/4 amount of work.
Take the top produces away, only 1/2 the production is done.
Now if those 300 shoplifters were all that could fit in store and 10,000 more possible shop lifters would shoplift but not able to fit inside, then remove 300 would only free space for 298 and Fat Billy in to shoplift. So, yes, then your power-law example would be correct, .. except if those remaing 10,000 shoplifters were migits with tiny hands that could not carry as much. Poor migits, don't throw them!
Some people are so dumb that prison doesn't deter them - and the answer to this is to not send these people to prison as much? Completely backward.
Really, it means some people cannot be trusted to not be removed from society. Like, do people not understand that *You can't do crimes when you're in prison*?
----------------------------------------------------------------------------------------
There seems to be a one to one overlap between YIMBYs and people who GROSSLY discount the social cost of crime. Which is really, really hilarious considering that crime ridden inner cities is an enormous part of why there is so much sprawl in America and why people don't want to live in the city as much as other parts of the world. Of course, YIMBYs are either too dumb or too ideologically blinded to put two and two together and would rather just call white people "racist" - apparently we have a moral obligation to put up with massive amounts of violent and property crime instead of trying to protect ourselves and our families from this at enormous personal cost.
Of course, they also assert that government programs will fix everything in spite of a completely lack of evidence for this beyond assuming every correlation that matches their ideology is 100% causation. Because hey, maybe if we give books to black families, black kids will start doing better in school, right?
In fairness, it's worth asking what methods best deter crime. Property crime seems to cycle with government food assistance (Meaning crime is higher when assistance has been less recent), suggesting that property crime is need driven. An average cost of a year in prison is about $38,000 to the state, with some states being much higher. This doesn't include the cost of law enforcement or prosecution, much less crime victimization.
On average, SNAP participants will receive an estimated $187 per month
https://www.cbpp.org/research/food-assistance/the-supplemental-nutrition-assistance-program-snap
So if giving SNAP to 17 people would prevent one person from being in prison for a year then it could easily be better to expand assistance rather than to increase prosecution for property crimes.
Of course, this is a very simplistic model that makes some taboo tradeoffs and doesn't account for feedback loops. Even so, it's worth at least asking: If we're going to spend public funds to reduce property crime, what's the best way to do it? While violent crime seems to correlate with socioeconomic status, short term economic assistance doesn't seem to deter violent crime, that I can see. (Improved nutrition does reduce violent crimes, especially when forced in controlled environments like prisons, but there seems to be a disconnect between increasing supplemental food assistance and improving nutrition.)
Matthew Yglesias, Official YIMBY Spokesperson + pariah on the left due to tough-on-crime, anti-woke, pro-private market views, would like a word. But don't let evidence stop you from righteously dunking on an outgroup, yeah? (Nevermind that Zvi is also YIMBY-pilled?)
Personally, I'm half-hoping for a "natural experiment" where...vigilante justice...is served to the minority responsible for All The Crime, and then suddenly all the whining stops because everyone finally realizes Crime Is Bad, Acktually. While I Do Not Condone This, if the government won't protect the people, then the people can't be faulted for protecting themselves...
Thanks for this. Just a note that when I studied theories of punishment in law school, the prof separated "deterrence" (penalties that prevent otherwise-might-be criminals from committing crime) from "separation" (penalties that remove convicted/repeat criminals from society, thus preventing them from committing more crime). I think it's a useful distinction instead of referring to both under the blanket name "deterrence."
Great work as always, Zvi.
On funding the police more - you also have to couple that with rules or laws that they can't just use it on traffic stops. I did a post analyzing this, and ~80% of police time goes to traffic stops and overhead, and this is while clearance rates for murder are ~50% and rape and assault <30% and property crime <10%.
Even if all we did was mandate "you have to spend 90% of your current "traffic ticket" time on actually solving crime," we'd roughly triple the amount of police hours put into actually solving crime.
The current allocation of police hours is absolutely bonkers relative to our clearance rates of serious crimes, and we should not be giving them more money while they're going to keep allocating time in bonkers ways.
Post here: https://performativebafflement.substack.com/p/more-than-80-of-police-hours-are?r=17hw9h
Don't traffic stops consistently uncover people with outstanding warrants, firearms violations, etc.?
Also, I'm under the impression that much of police time (especially overtime) spent doing traffic stops is funded by federal traffic safety grants.
> Don't traffic stops consistently uncover people with outstanding warrants, firearms violations, etc.?
If nearly all discretionary police time is going to traffic stops and closure rates are still <50% on everything important, just how many uncovered warrants and crimes can it be uncovering?
Clearly not enough, and spending time on actual police work might be in order.
I think there's something to what you say, but seems worth noting that "conduct traffic stops" and "investigate murders/rapes/assaults/thefts" are different skillsets, and presumably there are lots of cops who can do the former but not the latter.
Sure, but if we know they're wasting the vast majority of police hours while not solving any real crimes, it's a clear sign they're bad at allocating their resources, and we shouldn't give them any more without fixing that.
Obviously, do something like hire one detective instead of 2 or 3 traffic cops if it's a money issue, and we'll all be better off.
The world doesn't need 2-3 more traffic cops, that niche is filled and pointless anyways, but it does need 1 more detective who will actually solve crimes, and this writ large across all the cities will be a better end state for everyone.
I certainly would prefer a better allocation mix of fungible police man-hours and money (granting that they are, indeed, fully fungible). However, given how local traffic law enforcement cratered during covid and still hasn't reverted to the mean, I've gotta quibble with that pointless niche being filled. It's genuinely a worse equilibrium for me to live along a road that's now used for drag racing and other reckless driving on the regular! Some tickets are bullshit revenue-padding, sure, but it's not like moving violations aren't also crimes in their own right. (Perhaps a compromise would be diverting traffic cop money to red light cameras and other such surveillance? Probably more efficient anyway...)
I d prefer u wrote more about AI pls!
Dude, he writes about AI almost every day.
I anti-prefer this pls
"If we value a life at $12 million when calculating health care interventions, you’re telling me marginal murders only have a social cost of $9.4 million? That’s crazy, murder is considered much worse than other deaths and tends to happen to the young."
1. Not sure where you're getting these figures, but the $12 million figure might come from the value of a statistical life, which peaks around 46 years old, because it assigns the highest value to middle-aged adults based on economic productivity and willingness to pay for risk reduction. If you're using QALYs, ICER uses $100,000–$150,000 per QALY, so that could go as high as 10 million for a young victim? 9 million feels right.
2. But murder also tends to happen to people who are involved in criminal behaviour, and are more likely to be a net cost on society. For example, in this study, 80% of firearm homicide victims had an arrest record: https://icjia.illinois.gov/researchhub/articles/victim-offender-overlap-firearm-homicide-victims-with-and-without-criminal-records/. So I'm not sure I'd say murder is worse than other deaths.
It seems to me that the most constructive answer to "should we adjust prison sentences to be shorter / longer" is "mu" – don't focus attention on this question. As you say, the low hanging fruit clearly lies elsewhere (and there is likely a lot of it). We should focus on having a prison system that doesn't cause more crime after release, looking for ways for punishment to be more certain and swift, etc.
> None of this is easy.
I dunno, if you mean "easy" in the sense of "politically palatable" then I agree, it's quite a quagmire.
If you mean "easy" as in, "one can write a policy prescription that would work", then this seems pretty easy? Like, start with:
* Significantly increase police funding
* Possibly also increase court funding, if needed to keep up with police cases.
* In order to fund the above, either cut other parts of the justice system (e.g., prisons), or raise taxes, or both
* Get rid of police unions
* Bring back corporeal punishment
A harder thing to do, which doesn't seem that hard considering the currently shoddy state of things, is to improve the administration of police services. I'm thinking of things like,
* Identify the bottom quintile or so of police officers, and fire them.
* Figure out the right balance of number of officers, officer pay, equipment, training, etc.
One other thing based on the comments at Astral Codex Ten is that it's probably worth figuring out probation; the current system doesn't seem to work at all but I could imagine one that could work well at low cost.
Bottom quintile by what metric? We already have issues being caused by arrest quotas.
I did say it was “harder”. This is basically a management problem, and I don’t think the answer is quotas, nor probably anything too formulaic.
Two part suggestion;
1. for all but significant crimes then - Each local community would have a Judgment-Committee of 5 to 9 that are fathers or grandfathers that have raised children well that displays good prudent Judgment, and they would rule on local violations of good social behavior for adults 15-years and older. Such violations as defined to be against good social order, such as; lies, stealing, disrespecting social-standards, etc.
2. In today's society False-Witnessing, Lying, harming the reputation of others can not be corrected, so such abusers will never benefit in Loving Correction – a punishment to make clear they behaved badly and motivation to not repeat the abuse. Using Public Pain&Shame will pay the wages of their sin.
Punishment involves public pain and shame. Once a week, those to be punished would be taken to a public place like a park or public building court-yard and restrained to an anchor, and a collar attached to their neck, and for the directed period defined by the Judgment-Committee, the length of time that the collar delivers the taser sentence.
Those people that do not want to be punished need only to be respectful to those in their community and if they make a mistake they personally engage those they harmed and give satisfaction for the damage, if possible. If not, then the Judgment-Committee is used.
Those of age 14-years old would be required to view at least 3 of these punishments before turning 15-years old and be a risk of the same punishments for those community social crimes.
Repeat comment - but fits well here;
--
After reading about Nazi Germany's reEducation Camps
https://www.wintersonnenwende.com/scriptorium/english/archives/articles/ccfacts.html
and decided to use it a template to replace prisons. What do you think?
Modern Rehabilitation Camps: A Vision for Humane and Constructive Incarceration
Introduction
The modern prison system has long struggled to balance punishment with rehabilitation. In a society increasingly recognizing the value of human dignity and social integration, a new model of incarceration must emerge. This model, which we shall call "Rehabilitation Camps," draws inspiration from principles of order, discipline, and the promotion of virtue. Its purpose is not merely to confine but to reform and elevate the men within its walls, preparing them to rejoin society as honorable and productive citizens.
Principles of the Rehabilitation Camp
The core principle of the Rehabilitation Camp is to instill a deep sense of discipline, purpose, and moral virtue in its inmates. By emphasizing skill-building, social integration, and humane treatment, these camps aim to transform individuals who have fallen into criminality into dignified members of society. This system is not designed to humiliate or degrade but to uplift and inspire.
Moral and Social Education: Every inmate will be provided with a structured education in virtues such as honesty, self-discipline, accountability, and charity. Daily routines will involve lectures, readings, and discussions on moral philosophy, social order, and the responsibilities of citizenship.
Work and Trade Training: Inmates will be assigned meaningful and productive work according to their physical and intellectual capacities. This work will range from agricultural labor to skilled craftsmanship, preparing men for productive employment upon release. Employment training shall include apprenticeships, technical skills, and creative trades.
Physical Discipline and Health: Physical health will be seen as an essential aspect of moral clarity and personal strength. Inmates will engage in daily physical training to promote self-discipline and bodily health. Balanced, nutritious meals will be provided, and medical care will be readily available.
Guard as Mentor and Model: One of the most critical features of the Rehabilitation Camp is the role of the guards. The guards are not merely enforcers of rules but are to act as mentors and exemplars of moral virtue. Guards will be selected from among the finest citizens — men who embody self-restraint, honor, and compassion. Their conduct must reflect the highest standard of virtue so that they inspire respect and emulation from the inmates.
Gradual Reintegration and Parole: The ultimate goal of the camp is not indefinite detention but gradual reintegration. Inmates who demonstrate genuine reform will be granted increasing privileges and responsibilities, eventually transitioning to supervised parole and employment placements. The emphasis is always on reform and uplift, not punishment for its own sake.
Management of Violent and Impulse-Control Prisoners
A key challenge of any correctional facility is managing individuals prone to violence or severe impulsive behavior. The Rehabilitation Camp will approach this issue with particular care, ensuring the dignity and safety of all inmates while guiding the violent man toward self-mastery.
Controlled, Isolated Environments: Individuals with serious impulse-control problems will initially be housed in separate, highly monitored sections of the camp. These facilities will be designed not for punishment but for intensive correction. The goal is to protect these individuals from themselves and protect others from their aggression, allowing space for gradual reform.
Intensive Counseling and Discipline: Specialized counselors and chaplains will work closely with these inmates, using a combination of religious instruction, cognitive-behavioral therapy, and moral reasoning to assist the inmate in gaining self-mastery. Guards will be particularly chosen for their patience and virtue when working in these wings.
Physical Labor and Discipline: Labor has a transformative effect on unruly men. Those prone to violence will be given the most physically demanding but meaningful tasks, such as agricultural work, stone cutting, or timber work. The structure of hard physical labor coupled with moral instruction has proven historically effective in fostering discipline.
Strict Accountability with Dignity: Violent outbursts or breaches of discipline will be met with swift but humane consequences. Isolation or suspension of privileges will reinforce the importance of self-control. However, physical mistreatment or dehumanization will never be permitted. Dignity is paramount.
Pairing and Team Integration: Once a violent prisoner demonstrates a measurable level of self-control and discipline, he will be paired with another prisoner of similar progress to form a working team. These teams will engage in collaborative physical labor requiring cooperation, communication, and mutual reliance. This method serves to increase social aptitude and reduce isolationist tendencies.
Expedition and Autonomous Projects: As teams of two begin working efficiently together, they will be further integrated into larger teams of four. Once these four-man units demonstrate reliability, they will be assigned independent projects outside the main camp. One such project will involve leaving the camp for two to three days to construct their own sub-camp using provided tools and materials. The men will design and build their own cabin, utilizing creativity and teamwork. While largely free from guard oversight during these tasks, the men must adhere to clear task completion deadlines, dedicating three-quarters of their daily work hours to construction, farming, or resource gathering. This autonomous period fosters self-reliance and continued social skill development, preparing men for eventual reintegration into society.
Gradual Trust and Supervision Reduction: Successful completion of sub-camp projects will result in increased trust and privileges. Ultimately, men who demonstrate consistent reliability, moral reform, and teamwork aptitude will be eligible for early parole, transfer to open-labor programs, and eventual reintegration into society.
Case Study: Hypothetical Criminal in Two Systems
To demonstrate the potential benefits and drawbacks of Rehabilitation Camps compared to standard prisons, consider the case of John, a 32-year-old man convicted of armed bank robbery. John, a divorced father of a four-year-old son, faces a maximum sentence of four years.
Standard Prison Experience:
Social Isolation: In a standard prison, John would experience severe social disconnection. The limited family visitation and lack of community engagement would likely distance him further from his son.
Minimal Skill Acquisition: With limited access to meaningful labor or skill-building, John would likely leave prison without marketable skills, reducing his employability.
Psychological Deterioration: The punitive environment, lack of positive role models, and inherent exposure to criminal subculture could worsen John's behavior or mental state.
Limited Rehabilitation: Upon release, John would face immense difficulty reintegrating into society, struggling with stigma, lack of skills, and damaged familial bonds.
Rehabilitation Camp Experience:
Social Reintegration: John would work in teams, engage with mentors, and gradually restore his social aptitude, maintaining a sense of community.
Skill and Trade Development: Through labor assignments, John would learn practical skills in construction, farming, or other trades, improving his employment prospects post-release.
Positive Role Models: Guards and staff act as mentors, demonstrating discipline and virtue, which would inspire John to reform his own behavior.
Family Reconnection: Gradual reintegration processes, increased communication privileges, and visitation rights would help John maintain a relationship with his son.
Higher Rehabilitation Success: The structured, dignified environment of the camp increases the likelihood of John becoming a productive citizen post-release.
In conclusion, while standard prisons rely heavily on punishment and isolation, Rehabilitation Camps emphasize moral restoration, social reintegration, and skill acquisition. This model offers a clear path to societal contribution, reducing recidivism and restoring the dignity of incarcerated men. John’s case demonstrates how an approach grounded in virtue and productivity can transform not only individuals but also the communities they return to.
In this way, the modern prison system may evolve from a mere instrument of containment to a powerful engine of social and moral uplift, truly embodying justice and mercy in equal measure.
There are a few factors I'm surprised you didn't mention - though I'm sure this was already a lot of work.
Firstly, how much correlation is there between things that are (illegal + measured), and things that are (actually happening + harmful)? Eg, 327 professional shoplifters getting the catch-and-release treatment 20 times a year is obviously bad; but wage theft dwarfs all "conventional" forms of theft combined and is also punished by a slap on the wrist in the few cases it is even reported. A shoplifter may know they risk being arrested, inconvenienced for a day, and released; a manager committing wage theft knows perfectly well they will never be arrested and at worst the company they work for might be fined. Might it be more cost-effective to work on reducing wage theft?
Secondly and similarly, civil forfeiture laws have resulted in tens of billions of dollars being seized by police departments without due process, and the majority kept for their own use. Many of these seizures are never contested, because paying a lawyer to prove you *didn't* commit a crime costs more than the $2000 that was stolen. Some of this money is presumably taken from criminals, but to my knowledge we have no good way of knowing how much. This is a theft that is by definition committed only by police departments and could be much more readily stopped than chasing down individual criminals, trying, and imprisoning them.
I think this gets to the "victimless crime" distinction. Siphoning money is a widespread aggregate harm, and if we're just talking cents and dollars, it's [magnitude] more than the value of actual physical stolen goods. On the other hand, no one ever got shot during a botched wage theft, and there is indeed a qualitative difference when the stick-up kid wears white gloves, so to speak. Even just in terms of societal norms...Everybody Knows books get cooked, people fudge punches, veteran employees are mysteriously terminated right before qualifying for pensions. None of that erodes my faith in collective humanity, just individual managers and companies cashing out their integrity cheaply. But that's a different class of values than the idea that it's okay to just take stuff without paying, than believing that the average person would totally do a thieving if they could get away with it. Shitty for the employees too, I miss the days where I assumed the best of my customers by default and didn't constantly catch myself statistically discriminating. Power flows from the barrel of a gun, and all that...we don't actually want to go back to a world where what's yours is only so because you keep others from stealing it. So I think some amount of extra scrutiny/resource allocation is merited on the grounds of norm defense, above and beyond just the dollar value.
(Totally amenable to slapping wrists harder, of course! Chewing and walking gum! I just think it wouldn't be ideal at a direct 1:1 rate to trade e.g. shoplifting prosecution for wage theft prosecution. One must always solve for the equilibrium, as well...it seems pretty clear by now that recent spikes in shoplifting and other petty crimes were largely due to increasing awareness that there *isn't*, actually, much risk of getting arrested, let alone meaningfully punished. Paper tiger enforcement doesn't work so well once the cat's out of the bag. Considering how fast that spike was, I'm really cautious of ceding more such ground...it's so, so hard to recapture. Whereas wage theft is an endurance race, and the thought of seriously addressing it under the current administration anyway is...aspirational.)
Civil asset forfeiture is mostly bullshit though, yeah. Even to the extent it hurts actual criminals, it's like KYC-AML, where the squeeze far outweighs the juice...haven't found a single compelling defense of the practice. Yet it persists anyway, so...qui bono?
Ah, a glimpse into the AGI Normal timeline, where Zvi blogs about "mundane" topics...
Timely post to read after a 5/10 harrowing late-night train ride home with a crazy high guy (really high, or high as well as crazy? guess it doesn't matter)...and the whole time I'm thinking, ugh, I just want to decompress on my commute home and not worry about shit like this, can we not, why is it so hard to deal with even the Obvious Ones. That's the social costs angle: even though nothing serious actually happened, there's clearly a lot of counterfactual value being set on fire, for everyone raising guards on the train + waiting on platform. Big Shut Up and Multiply energy. I'd pay a lot/have to be paid a lot to (not) be assaulted. And even more to go back to societal norms of mostly not having to worry about any such interlocution at all, of course...which wasn't even that long ago! ~Six years is within living memory of just about everyone!
Your methodology note is interesting...does it vary by question, or is it a more general shift? Criminology is well-known to be riddled with shitty epistemics and confounds (hell, even getting reliable data on actual non-murder crimes is hard), so I can understand being extra skeptical of the classic More Than You Wanted To Know treatment. Academic GIGO is still garbage, there are alternative ways to accurately map the territory. But some of your other notable "takes" posts like Contra Deceptive Car Seats, Jonestown Act Massacre, Never Sports Bet Against A Sicilian, etc. do rely on studying the studies? I guess the treatment is different even when working from the same epistemic base though...still enjoy Scott's analyses, but he seems (increasingly?) too nervous about rolling to disbelieve, deriving from first principles, etc. Admirable approach, falls short when the Official Data just isn't there/too contradictory. It's hard to live one's life on responsibly uncertain hedge bets, sadly, especially for immediately-relevant issues like crime and punishment that are actively being contested.
hey Avalanche,
Have your sons join the HS westling team, and your daughters take up the batton and gymnastics. The batton handing skills will unknowing turn her into a deadly weapon handler .. I know my older sister learned the batton and darn hear knocked me unconsious not realizing how serious a blow she game me.
And if your sons get confortable in public battles with strange young men his size and if he wons most often, your sons will make soild eye contanct with the psychos on subway or the gang of three and they will know they should target some one that woundn't maybe F-them-up and then when they down stomp on they necks so they never ever target anyone else - back to Dust with You!
But tell your sons when they go to keg parties with pot to not get into any fights in parken lot and thow him around like rag-doll bounding him off the cars - because he might not hear Holy Mary's warning that he is about to kill a man, and that is Forever.
Hope this helps your [future] children.
Podcast episode for this post
https://open.substack.com/pub/dwatvpodcast/p/response-to-scott-alexander-on-imprisonment