I suspect many of the opponents of the use of facial recognition technology (but perhaps not all of them) are not opposed to the enforcement of criminal codes,but rather very skeptical that governments can be trusted with such tools. Considering the already highly intrusive surveillance state apparatus that exists in the five eyes countries - the thought of adding in facial recognition technology is quit terrifying.
With cameras everywhere this could give you a capacity to track where everyone is at any moment - especially if everyone is installing doorbell cameras etc whose corporate suppliers are likely to bend to rubber stamped warrants or make deals with the NSA to further government surveillance.
There's some of that, but there is also an "enforcing the law is bad" crowd. There was an Atlantic article a few years ago that was aghast at the fact that Ring as were resulting in people being arrested for stealing packages.
That was a long and wild read. I did have a caveat in my original comment, in anticipation that some people might actually hold the view that MY discussed. I have certainly heard about progressive jurisdictions where theft just isn't prosecuted, but the whole concept seems so crazy it's still a bit hard to believe.
Yeah, it's nuts. I understand that there are reasons why police might not go all "Inspector Javert" over a couple stolen packages, and that we may not even want them to do so, but the thesis of this article is very much that it is wrong and, perhaps racist, for people to get upset that someone is repeatedly stealing their Amazon packages.
I suspect that a regime that is going to illegally abuse facial recognition is also not going to be stopped by a law against using facial recognition at all. We might as well have it as a tool for legitimate law enforcement investigation because I doubt that some "ban" on it as a technology is really going to save us from Big Brother.
Right. The slippery slope argument is bogus. If there's any autocracy in America's future, that regime is going to use whatever tools technology makes available. And there won't be any debate.
Those arguing against greater use of these identifying technologies need to point out why rule-of-law constrained governments and police shouldn't use such technologies. There are arguments that can be marshalled, for sure. But "it puts us on the path to Big Brother" doesn't seem like one. What puts us on the path to Big Brother is things like elections nullification, and subversion of democratic norms.
Keeping it illegal does afford greater protections for citizens than not - for example you could claim fruit of the poisonous tree, to get evidence excluded where it was derived from the illegal use of facial recognition mass surveillance. Similarly, it limits options for funding and deploying the technology.
Also: if we rely on face recognition, we should make sure that the judicial system is able to find exculpatory evidence on this basis: if you're accused of a certain crime at a certain time and place, your defensive team should be able to look for evidence of your presence elsewhere (which is not easy as recordings are often deleted for reasons of privacy and storage capacity).
You are perhaps over-estimating how expensive this software is, or at least how expensive it will be in the not too distant future. More than likely, if an evil repressive government wants this technology it's not going to be hard to get it.
Agreed. Let’s not kid ourselves that the database wouldn’t also be full of Trump supporters merely attending a rally or abortion activists marching in protest. This information could easily be used to harass opponents of the ruling political party, justifying searches and detainment just because having a match in the database provides cause.
Either you have a society where the government is empowered to harrass people for their political views or you do not. The Soviet Union was quite the archetypal totalitarian state without a lot of high tech tools. I don’t think today’s China has anything on 1950’s Russia in that regard.
>>Agreed. Let’s not kid ourselves that the database wouldn’t also be full of Trump supporters merely attending a rally or abortion activists marching in protest.<<
I don't think we need to "kid ourselves" this wouldn't be the case. We need to pass laws and formulate regulations preventing this from being the case. If we can successfully do this, why not opt for less crime? Are you claiming we cannot do this? Perhaps access to the database would be provided only upon issuance of a warrant...
There are things that people want to do that are ethical but they don't want the whole world to know about it.
Maybe they are into some consensual but potentially not-well-accepted sexual interests. Maybe they go to AA meetings or a psychologist. Maybe they are involved in a controversial political group. Maybe they are in an insular religious community but are considering leaving. I can think of millions of reasons why someone wouldn't want all of their daily whereabouts monitored.
We don't want to live in a world where the government has so much surveillance information that they can easily harass, intimidate, or blackmail someone that they don't like with this information. So if we are going to increase survaillance for better enforcement of criminal laws, we need assurances that these measures will be only used for investigating crimes, and that authority to access the data is highly limited both in terms of who has access and how the data can be used.
Who is going to look at the results of this data from every camera on every corner in every city and town in the US?
There should be regulations that confine analyzing this data, that doing so without a clear reason is unacceptable (cause for dismissal.) That's the case for hospital employees; you cannot look up a patient's information without a good reason.
My reading was that Matt was advocating for *more* cameras, which I assumed would be government supported. I don't imagine people would watch this footage for casual fun; they would be looking for a criminal. Therefore anyone who does not seem to be a criminal would not be subject to facial recognition technology, which I guess is the basis for people's objections. Not sure if they're worried about being put in some kind of national facial recognition database for future punishment/ oppression, or about giving criminals an unfair disadvantage?
I do think that there's a difference between a world where there are cameras in some public places, with the data only being able to be parsed by a human is different from a world where cameras in public are nearly everywhere, and your face can be automatically recognized and all the data can be automatically processed, analyzed, integrated and stored at a large scale. Perhaps the latter world requires some reanalysis of the pre-existing social contract
And just to add it shows society does a pretty good job ***not*** sharing this info.
You almost never see videos of regular joes blowing their life saving in slot machines for 8 hour shifts even though the videos exist. Nor do you see videos of guys often slurring their words. Occassionally some pure schmuck gets exposed. But by and large we never see this stuff.
This is a good point, but the kicker is less about the government having technology like this and more about the fact that the technology itself is extremely opaque and cops, attorneys, and juries are all going to be equally unable to engage with it productively. Law enforcement is just not equipped to deal with this.
How do we monitor the algorithm? Who gets to pressure test it and make sure that it is doing what it says it's doing? What happens when version 3.4.96 identifies your kid in the grainy screencap of someone getting out of a car, and version 3.4.97 disagrees?
I strongly suspect that juries will just accept "computer says it was you" as an extremely compelling argument, because that's how most people interact with technology.
I work in ML/AI and volunteer with a group that encourages BIPOC young adults to learn about and pursue careers in math/science/technology (I mentor in ML/AI/scientific computing). What you're saying is exactly what someone in the know should be worried about with facial recognition. It's not that good, it's lazily opaque to the average citizen, and it's accuracy diverges significantly by the race, gender, and age of the person being classified. Guess which groups it performs the worst for -> https://sitn.hms.harvard.edu/flash/2020/racial-discrimination-in-face-recognition-technology/
I don't think the main point of the technology is to decide who is who on a video once it gets to court. I think the idea is more "a car was stolen yesterday, we have video showing random dude doing it, we have no idea who he is, lets run the picture through a database to see if a name pops up" at which point you use that as a lead to find them and then use the video itself as evidence and let the jury decide how much the accused looks like the person on the tape. It's like a faster and more certain version of releasing the video to the media in hopes some viewer recognizes them.
Yeah, I think we might disagree here. I think that the scenario you describe sounds plausible, but note the basic circularity of the logic that you are using:
There is a video of someone that committed a crime, and the cops search a large database to find a person that looks like the person in the video. The database is large, so the computer works hard and finds someone (or probably a lot of people) who plausibly look like the person in the video. Then the cops present the video as evidence that the person that they arrested committed the crime, and asks the jury to decide for themselves whether the person on the stand looks like the person in the video. And he probably does, because the computer is pretty good at finding people that look like that. But did the computer pick the right person among all of the people who plausibly look like the video? Is the right person even in the database to begin with? There's not really any way of knowing without secondary evidence.
In a lot of cases the cops might find a lot of secondary evidence when they investigate whomever the computer selects, which is fine. But in a lot of cases they won't, and in cases like that I'll bet there would be a real temptation to use the exact same logic that you did above to get a conviction anyway.
Fundamentally though, how is that a more likely result than a similar misidentification from broadcasting the video on the local news and having someone call in and say "that looks like my co-worker!" The technology is really just another tool to find the suspect and at the end of the day the plausibility of that being the person in question has the same burden of proof as it would using contemporary methods.
Again we might disagree, but I think it's basically because in practice people give a huge amount of deference to technical solutions because the underlying issues are too abstract to reason with.
If a person gives a tip to the cops the defendant's attorney can cross examine the informant and ask a bunch of pertinent questions to clarify their basis for identifying the person in the picture, which amounts to additional evidence supporting or disputing the ID. None of that secondary evaluation of the ID itself is really available in a database search, because an attorney can't ask the convolutional neural network why particular layers were tuned the way that they were.
So put up guardrails and trust that the USA is fundamentally different than China.
The biggest factor to unsolved crime is the "street code". If there's a crime and no one in the neighborhood wants to talk, put up cameras as a last resort. This kind of thing could go a long way to reducing inner city violence.
This is an intuition I hold, too. Laws on the books we're not going to enforce are actively harmful. The Anker paper backs this intuition up from another direction.
This is my hang-up. I agree with most of what Matt's said above, but giving the government the tools to track its citinzenry's every move... Let's just say there's a reason China is the main exporter of this technology/civil model.
I broadly agree , but I think this piece’s main weakness is that it’s written against the backdrop of Twitter crazies who oppose law enforcement. Thus it doesn’t address the concerns of normal people who instead prioritize the well being of the average, law abiding citizen. Us normies are all for catching (and punishing !) criminals , but we have other concerns about expanding police powers. How do you make sure innocent people are not falsely accused? How do you balance these real needs with privacy concerns (e.g. do we allow any and every cop unmonitored access to the most powerful facial recognition tech?). In short, in a healthy democracy we should always take pause before giving the government new powers. Ultimately, as often, we may conclude with a sigh that giving this power is a necessary evil, but we may want to think about adding some checks for reassurance. I wish MY would have addressed these mainstream concerns.
I can't speak for every jurisdiction, but in the ones I am familiar with, courts absolutely are aware of the low reliability of witness identifications, and the rules of evidence have been re-written to address and mitigate these concerns.
I hadn’t heard that, but if it’s right it’s going to lead to even fewer criminals being punished unless we can make up the gap with more reliable tools.
eye witness testimony requires less lawyerly skill to introduce at trial than any other form of evidence. rules against hearsay make it difficult to testify to what a witness learned from others. what they personally saw is fair game unless it is completely unconnected to the charged offense.
one can certainly call an expert witness to impeach identification testimony, but few criminal defendants can afford experts. i have one client who has a very good job and she is struggling to pay for a single expert. she’s also one of my better off clients.
Can you vaguely indicate which jurisdictions these are? I'm an NA lawyer and while we're vaguely aware of the sketch nature of eyewitness evidence at least my jurisdiction still heavily favours it over other forms of evidence.
No, but we do have strict rules about evidence admissibility, perjury laws, the critical role of cross examination etc. centuries and millennia of experience taught us how to use witness testimony correctly and beware its pitfalls. We should aim for no less when incorporating new methods.
IMO the Federal Rules of Evidence and their state analogues are not by any means paragons of the correct application of evidence (inter alia they're at least in part explicitly counter-Bayesian) and reflect the staunch reliance on tradition to substitute for considered and principled reasoning. There's a prominent, and in my view unfortunate, strain of thought that essentially treats Anglo-American evidentiary norms as good by dint of their persistence and pedigree rather than because they bear the slightest relation to how actual humans conduct themselves and make assessments outside the confines of a court of law.
Not everything in them is bad, mind you, but treating them as the considered result of decades and centuries of evolutionary pressure to reach the correct result is a long way from the correct paradigm in my view.
omg yes. on the other hand, the first thing clients want to do is dig up/drag in dirt on their accusers. the instinct is very strong and transcends race, gender and class. blah blah blah uses drugs, had an affair, whatever, so predominates what i hear it can be hard to focus clients on their actual cases.
also, the legal definition of relevance (rule 401 i think) is very very broad. the principle that probative value just outweigh potential for unfair prejudice is 1) obvious and 2) too i’ll defined to be much more than a blank check made out to trial judges
In the one case where I was selected for the jury, I was shocked to see that the police officer who testified was not allowed to read from his notes while testifying, but was allowed to look at his notes to "refresh his memory" and then testify again after doing that. I can see that there's some intuitive idea of how the memory works on which this set of restrictions and permissions might make sense, but not on any realistic idea of how it works!
If you have JSTOR access, I think this paper by Laurence Tribe does a remarkably good job of arguing that there are certain ways the standards of evidence *should* be anti-Bayesian: https://www.jstor.org/stable/1339610
The biggest and most prominent examples are FRE 404(a)(1) (generally referred to as "conduct in conformity" [with the defendant's character]) and FRE 404(b)(1) (disallowing prior bad acts as evidence of a defendant's character). https://www.law.cornell.edu/rules/fre/rule_404
(a) Character Evidence.
(1) Prohibited Uses. Evidence of a person’s character or character trait is not admissible to prove that on a particular occasion the person acted in accordance with the character or trait.
[...]
(b) Other Crimes, Wrongs, or Acts.
(1) Prohibited Uses. Evidence of any other crime, wrong, or act is not admissible to prove a person’s character in order to show that on a particular occasion the person acted in accordance with the character.
Separately, the hearsay rules more or less in their entirety. In brief: the world you and I know and live in and all of private enterprise basically run on hearsay (modulo the fact that technically hearsay only exists outside court of law and thus literally everything outside a court is hearsay), and in courts of law it's presumptively disallowed.
I think this is all kind of addressed, at least implicitly. The thesis is that more effective policing means less crime means fewer accusations at all, which should lead to fewer false accusations. Better technology use should lead to fewer false accusations and/or easier exonerations (as we’ve seen with the expanded use of DNA evidence).
I get the apprehension about giving the police more tools, but I think you need to make a case of what they’re going to use those tools for that’s different and worse from what they already do. I think that’s an easy to make when the new tools are military weapons and vehicles. With surveillance stuff? I’m not sure how much would change given the degree to which we’re already tracked by cameras and internet stuff. What would they use it for that they can’t do already with eyewitnesses, patchwork video availability, etc; and why would it be bad?
IMO it’s addressed only under the implicit premise of a police force that’s 100% professional and in good faith. I’m far from those demonizing the police, but let’s not forget that they’re human, that power corrupts etc and frankly that beyond the theoretical concern evidence isn’t lacking for abuse of current police powers by bad actors on all levels. I think these concern deserved explicit and comprehensive attention in the piece.
But what part of this involves giving the police more power or relies on them being 100% professional and in good faith? They can already arrest whoever they want with lower standards of evidence. They can stop and kill people with relative impunity. The people that they do investigate, charge, and convict end up in prison for a very long time! IMO one of the underlying assumptions here is that the current police aren't actually that good in the US so we need to find a way to make them better. So let them use facial recognition instead of relying on eye-witnesses and racial profiling. Use ankle bracelets and breathalyzers so there's less differential ("racially problematic") enforcement of bail/probation conditions. Speed cameras instead of DWB! (I know that last one is from a different piece than here, but it's in the same spirit).
Assume that the police are... sub-optimal (I certainly do!). But then tell me how these proposed powers (and I'm not sure how much of this I would classify as new powers per se, but ok) would lead to them doing worse things. As I mentioned above, this is easy to do for the military stuff: If you give cops more guns they will shoot more people. If you give them tanks and shit they will act more like an occupying force with any provocation (including riots or whatever, but also even just peaceful protests). Even if you just give them cool tactical gear (helmets, vests, cool belts, etc), it will encourage the cosplay warrior crap which makes them less effective and more dangerous to the public.
Can you make a similar case against using facial recognition and location monitoring? I know that sometimes people use "don't break the law if you don't want to be punished" in bad faith. But I think the point here is that if you have effective crime detection, then you won't get falsely accused of breaking into houses if the GPS shows that you weren't breaking into houses. Or if you don't drink above a certain amount then the cops won't waste time following and harassing you to check up on whether you're complying with your probation/parole conditions. So what is the case for "using these technologies would let the cops do new bad things" instead of doing fewer of the old bad things?
To be fair, if you're just of the opinion that they'll find a way to abuse anything given the chance, I 100% won't argue with that. But I do think that there are levels of degree, and that what's proposed might lead to less police misbehavior than the status quo.
If police officers are too "fucking stupid" to use valuable tools to reduce the crime rate the solution is to recruit smarter people into policing not ban the use of valuable tools.
This doesn't make sense to me as an argument against more non human surveillance. If you don't trust the police, wouldn't you want there to be evidence outside of a detective saying "he did it?" If you a juror, would you feel more confident convicting based on eye witness testimony or on camera/DNA evidence?
Neither, or either . I’d be confident convicting based on the weight of the accumulation of the evidence , the more the better, but mostly the better scrutinized the better. Camera/dna evidence isn’t some objective truth immune from heuristic challenges and manipulation and it requires rigorous methodologies of application just like witness testimony.
In any case all this is a separate issue from the question of abuse. The answer to that is simple. It’s not black or white. You don’t write someone a blank check but you do trust them with a signed check for a specific sum despite that also compromising you to an extent (your routing and account numbers are on there etc). Life requires balance. With government we need to try to get good people there but also put guardrails on them to guide them to do their work properly and not abuse their powers.
I suspect that a good deal of the concern about facial recognition software is rooted in a distrust of the police and how they will abuse it.
Other than senior citizens whose opinions are shaped by television police procedurals, a lot of people (and not just Twitter crazies) believe that police are prone to abusing their authority -- particularly when it comes to interactions with Black people.
Because of that distrust there is probably a stronger than is actually warranted reluctance to give police even more tools that could be abused.
I think the trust/distrust dichotomy isn’t helpful. There are very few angels or devils out there. We need to rather think in terms of cost benefit analysis (as MY does) but then *also* account for potential abuses both within and outside the legal procedure and regulate to minimize them (the issue MY overlooks).
P.S. I strongly recommend reading John McWhorter on the police and black people. Some of the worse things most liberals now take for granted are simply factual untrue. It’s a very sad state of affairs.
"this piece’s main weakness is that it’s written against the backdrop of Twitter crazies who oppose law enforcement."
I don't think the idea remains with just "Twitter crazies." In my undergrad, that ethos infused many anthropology/sociology/ethnology/South Asian studies talks, because the professors in those disciplines regularly worked in non-democratic or authoritarian countries for which the legal system did not accurately reflect the practices of the populace, and greater enforcement means greater hardship. I suspect the principle then diffuses out to people who study the US legal system, without anybody really paying attention to its preconditions for relevance (also relevant: Matt's discussion of trans-Atlantic pollination in last week's mailbag).
But yeah, in a democratic society, the concept doesn't really make sense.
I never meant to imply that Twitter crazies don’t have day jobs, but that’s not how they’re influencing MY and skewing the emphases in his public writing.
I think we need to talk about the elephant in the room: if catching criminals is good, it will involve catching a disproportionate number of black men.
Progressives don't like enforcing crime because it will target black people much more commonly. If we use facial recognition software, or make more extensive use of DNA databases, they will be used much more commonly against black men. Of course, if these tools ultimately reduce mass incarceration while reducing crime, then everyone is better off, including the black men who would ultimately end up going to prison. But progressives will fight these measure tooth and nail. Expect articles in Vox about racially biased facial recognition, for example.
Liberals would be better off talking about this stuff, just as conservatives would be better off talking about a coherent health care policy.
There’s no elephant, and this fact was always well known. The other side of the coin is that black communities will disproportionately benefit because they are disproportionately victimized by crime. Such policy would thus be neither racist nor “affirmative action” but simply the state doing its job to protect its citizens. The problem with disproportionate black crime has root causes that need to be addressed as such but the idea that giving black criminals a pass on racial grounds is a legitimate or helpful measure let alone some “anti racist” necessity is pure insanity that thus never occurred to anyone before the last decade, and it’s still totally unclear to me how this toxic nonsense ever came to mainstream attention.
I think you underestimate the elephant content of the room. I've encountered progressives in real life, not just on Twitter, who are genuinely willing to take the position that wealthy white people commit *violent felonies* at least at the same rate, or possibly even higher rates(!), than poor black people, and that the reason that's completely unreflected in crime statistics is that police and/or prosecutors cover up this wave of white criminality, if it ever gets reported in the first place (which they believe in many cases it isn't).
Someone made this argument to me on Reddit the other day! They added the additional theory that rich white neighborhoods were intentionally under-policed for this exact reason.
Yep, I've encountered that variation in real life too. It typically comes up when someone is explaining how it's racist that there's so much more police activity in [POOR NEIGHBORHOOD] than [RICH NEIGHBORHOOD] and I respond by pointing out that [POOR NEIGHBORHOOD] has dramatically higher violent crime rates, to which the other person replies that [RICH NEIGHBORHOOD] is really just as violent, but the police stay away from it because it's full of rich white people. If I proceed further and question why rich white people would tolerate all that crime in their neighborhood, I usually get a response that sounds like some variation of "Get Out"/"The Purge" where the rich white people are supposedly beating/raping/torturing/murdering non-white workers or random passers-through.
My experience of the "black community" indicates that today it's still very much the same community which backed Clinton's '94 crime bill to the hilt.
The hipster lefties just want to pretend that the 12 black humanities Ph.D grads that they follow on Twitter are "representative" of African Americans, of whom under a quarter have a bachelor's or graduate degree, and have erected a circular firing squad and peer pressure architecture around that stupid shibboleth.
I've always thought that there's a niche waiting to be filled in The Discourse for a Black commentator pushing back against those 12 Black humanities Ph.D grads. There's probably already someone out there that I'm just not aware of. Someone like Ruben Gallego telling people to cut it out with the Latinx crap.
Roland Fryer, Thomas Chatterton Williams, Kmele Foster, Glenn Loury, Coleman Hughes, even Adolph Reed. There’s no shortage, but our gatekeepers on the liberal side seem to have decided that they’re all suspect somehow.
No ethnic or racial grouping within the US is a monolith.
But I regard it as a particularly stupid exercise for the woke/hipster left to anoint a very small, skewed sample as representative of such a "community."
“They are disproportionately victimized by crime” really is something that we don’t seem to talk about like we should, and I find it heartbreaking. The vast majority of people in zip codes where bullets fly frequently are law abiding citizens. Think about the way we talk about school shootings, and think about the way we *don’t* talk about being a person trying to raise a family in those neighborhoods. It’s horrifying. School shootings terrify us and dominate conversation even though they’re statistically rare. There are places in this country where you really do need to worry about shootings, not a statistical rarity at all, and we (those of us who don’t and won’t live in those places) just sort of take it for granted.
The people in that situation have less representation in our national discussion about crime than the criminals do. If white people with money had to live that way, this country would already be fascist because if it took a police state to stop it, we’d have a police state.
It’s like we’re so afraid that some conservative will bring up “black on black crime” that we make little or no effort at helping the people who are being terrorized. Again: the vast majority of people who live in our nation’s scariest zip codes are 1) black and 2) law abiding citizens. How is this not one of the most pressing concerns of people who care about equity?
Also, given that perpetrators of black on black crime are less likely to be caught than perpetrators of black on white crime, and given that most murders are intra racial, more enforcement will mean more black people going to prison
If punishment is like lightning striking, if the randomness of punishment blunts deterrence, and if a lot of black people are being struck by lightning, a sensible anti-racist might want lightning strikes to be gentler.
“Like lightning striking” presumes both random and rare. If something happens all the time, it’s no longer “random” especially when it’s happening to people who are committing the same kind of action.
Right now there are far more white people getting away with crimes than black people (see my comment on another thread about drunk driving), so it’s much more likely that catching more criminals will probably result in the cohort of caught criminals having a racial makeup that is closer to that of the rest of the country.
I think the important thing to note is that basic moral decency requires us not to give half a fuck whether this is true or not.
Enforce the law, protect the innocent, and send the guilty to prison, and if we do it well enough we will probably find that there are fewer guilty people and we can probably imprison them for shorter terms.
You assume there's one offender per victim. What if one group sees a higher rate of repeat offenders than the other? (I do not know the answer to this question and any guess on my part would be pure conjecture, but it's a possible explanation for the disconnect.)
Bargain with whom? Whose interest precisely is it that *any* kind of crime be under-enforced that it should be a “bargain” to enforce one or the other ?
The claim is that it would be unfair to step up enforcement of the laws that marginalized people break, while white collar criminals and corporate wage thieves go unpunished. Sometimes extended to claim that because law-and-order sentiment focuses on marginalized-people crimes, it is in fact just animosity towards marginalized people and it a commitment to the rule of law in itself.
How on earth is it unfair to enforce the law? Lack of enforcement among marginalized groups is a particularly nasty social ill and injustice. I’d go as far as to say that the pro-crime (pseudo)-leftists are engaging in class warfare against the poor (who are— as we are obliged to recite nowadays—disproportionately poc)
To be fair, there has been lots of squawking from Republicans about the increase in enforcement personnel for the IRS, which is kinda analogous to the "defund the police" silliness from the left.
I don't think it's unfair to enforce the law; that's why I'm interested in political strategies that could make it happen. But the "pro crime" leftists are an important part of the electorate in a big city, and to get anything done you're going to have to address their fundamental concerns, like the (not unfounded) complaint that law-and-order, policing, and incarceration seem to be more about suppressing the poor than something also binding on the rich.
There is a great deal of evidence that “marginalized people” commit crimes far out of proportion to their population.
Is there any evidence whatsoever that there are appreciable numbers of white collar criminals or “corporate wage thieves” (whatever those are) are breaking the law with impunity?
My understanding is that "corporate wage thieves" are when people on hourly wages are forced (by threat of losing their job, not physical coercion) to do extra work "off the clock" for which they are not properly compensated. The comparison I once saw, which always strikes home to me, is that if a worker pulls $100 out of the cash register, puts it in his pocket, and walks away then that is considered a serious crime and he will be arrested. If an employer decides to simply not count $100 of a worker's labor and leave it out of his paycheck, the police usually consider that to be an employment dispute and not a criminal matter. Yet in both cases $100 was stolen.
I've seen that explanation before, but I tend to find it uncompelling because most, if not all, states have plaintiff friendly anti-wage theft laws to encourage civil enforcement (e.g., statutory minimum damages, doubling or trebling of the stolen amount, attorney fees, etc.) and which are certainly enough to make wage theft a poor business strategy.
Ok, I may have phrased it badly, but my point is rather the “grand bargain” framing as problematic. We all know that the republicans are traditionally there to help big money. But the phrasing here suggests the democrats are there to help violent criminals. That’s rather a rough rebuke.
Ken, what company to you work for, what industry is it in and what are you doing to help your state enforce their labor laws to level your playing field? Also, what are you doing to keep undocumented labor from undercutting documented labor? Probably nothing.
There’s an interesting tension in Matt’s thinking. He has talked about keeping things like prostitution and cocaine formally illegal but not enforcing such laws vigorously. The idea is to stigmatize and reduce such conduct without oodles of caging. Yet he also argues that swift and likely punishment is the best deterrent.
The basic problem is that a system designed to stamp out murder and robbery has been pressed into service to reduce vice crimes. This means that resources devoted to vice suppression inevitably come out of appropriations and manpower tasked to prevent violent crimes. At a typical felony arraignment calendar in Georgia, a third of the people are accused of victimless crimes. That’s a real problem when getting a fairly simple armed robbery case to trial takes over a year.
Punishing vice through civil penalties (possibly a portion of the offenders wealth/income) would be much better
Or just involuntarily commit addicts and people with debilitating mental illness to rehab and treatment. How is giving a large fine to a heroine addict with no money going to improve their problem?
Portugal's system for substance abuse seems pretty good to me. We usually talk about things on a punitive / rehabilitative axis and looking through that lens it's probably a bit more punitive than many Americans imagine, but certainly rehablitative overall.
But to me the major difference in focus is accountability. Portugal's drug courts expect you to not abuse illegal drugs. They'll assist you in stopping if they can assist you, but if you continue to resist they'll increasingly use the stick alongside the carrot.
American policymakers seem to come at it from either a "more stick" or "more carrot" angle. I think we'd be better off thinking of it as "more accountability / more expectations". I wouldn't accept a family member of mine using heroin, for example, because I love them. I expect them to do better. We should love our fellow citizens enough to expect them not to abuse drugs.
Committing someone to rehab without them actually wanting to rehabilitate is pointless. All your really doing is forcing them into a different type of prison. That's a big step to me and deserves a noticeable hurdle for society to impose upon an individual.
The first two clauses of this seem like they're driven by the moral view expressed in the third rather than any sort of empiricism, given the data we have globally.
Also, as Wigan pointed out, there's a court system for this in Portugal and countries which follow that model. There *is* due process.
Making mental health treatment more widely available (like actual inpatient treatment for more than 10 days, and follow-up instead of just tossing patients out with a bunch of prescriptions) that might also prevent a lot of crime. A large number of drug addicts may be “self medicating” conditions that could be controlled with professional treatment.
I really did mean substance addicts. Those people should be priority for enforcement of "vice" crimes. The lawyer who does a line of cocaine every few months isn't high priority.
Sure - but also, don't most significant substance addicts start off as occasional users who ramp up over time into what you see on the streets. So if we made people think that doing the initial dose was likely going to be caught and actually penalized then wouldn't it decrease the number of substance addicts over time?
But how likely are they to actually pay civil penalties? (I don't honestly know, my priors are "rather low," but not very strongly held when it comes to specifically issued citations, summons or tickets. Don't have any stats ready to hand).
30 million Americans are debilitated (meaning seriously impaired in their ability to function in daily life) by severe mental illness and substance addiction? Seems high.
Overall I agree with your approach. Less harsh sentences combined with a higher likelihood of getting caught if you commit a crime is the right way to go.
I think that increased survaillance is a reasonable punishment for committing a crime (and less cruel punishment than incarceration). However, I think it's worth discussing limits of surveillance. People who don't commit crimes rightly don't want to live in a surveillance state (like China) and are rightly concerned about issues of privacy and government control of daily life. So where do we draw the line? (I ask this as a genuine question, not a rhetorical one...as with most things, I'm sure there's a reasonable line to be drawn, and that those who use it as a rhetorical question usually are suffering from a slippery slope fallacy).
I was disturbed that the South Dakota program required abstinence from people on bond for dui who had not yet been convicted.
I’m also not sure that abstinence should be required in every first time drunk driving case. If there is an accident or the BAC is over 0.12 or there is a prior conviction, sure, but I don’t think someone driving home from a dinner with wine with a BAC of 0.09 should be subjected to the surveillance state.
I'm my personal experience .09 is actually like, pretty horribly intoxicated, but that gets to the real issue, which is that BAC is actually a trash proxy for impairment. Depending on the person, their tolerance, what they drank, when they drank it, what else they ate, their genetic predisposition, etc etc. Some people are sometimes trashed at like .04 and others are fine at .12. The breathalyzer system is really almost indefensible, even more so given how many of the impaired drivers out there are impaired by things other than alcohol.
How does it violate the 5th if its a condition of parole? You're free to say no to using it, but then you stay in prison and serve out your sentence...
Mandatory breath tests do violate article i, section i, paragraph xvi of the Georgia constitution, but driving is a state issued license, not a right, so a license can be conditioned upon taking a breath test when there is probable cause.
The constitutional amendment is basically dead. But if SCOTUS declared it a violation, I think you could get a breathalyzer exemption passed very quickly.
I'd imagine allowing breathalyzers would cross 95%+ in support. I spent time in a public defender's office and I'd support it.
Opposition would be limited to an extreme minority of people with very strong priors on restricting police power.
On international standards, a cutoff of 0.08 is on the high side. Where I live now has a cutoff of 0.05, and cutoffs of 0.03 or 0.05 are not uncommon elsewhere. (In Brazil, they allow no alcohol in your system...)
Anyway, I am fine with treating seriously driving at 0.09!
I've seen people argue that we should lower what BAC is unacceptable, but also make it a sliding scale of punishment. I.e. a .04-.07 is just a fine and you're not allowed to drive for the night. All the way up to, a .15 is an actual felony and not just a simple DUI. That way we can gradate the crime for how bad it actually is. If you're a .05 that's about as 'impaired' as driving sober but sleepy, which I think all of us have done at some point
Lol, I think maybe you're in the wrong place? Half this comment section are transit fanboyz/girlz, MY is widely known for his pro-transit takes, an the lone avidly pro-car guy gets flack all the time.
Just last week I said, "I want better transit partly so we can take away licenses for all sorts of offenses and people don't get to bitch" or something along those lines.
Yea, the problem with .08 is that it's basically a way of saying, "we want to charge people for reckless endangerment, but that's hard, so we'd rather implement a largely arbitrary cutoff that's easy to enforce in court but is actually poorly related to the thing we're actually trying to get at."
The problem with something like the .02 concept is that it it's largely unenforceable except as a post facto enhancer of other infractions because there's no outward evidence of impairment. It's also hard to treat "DUI" as a serious infraction when you're talking an .02-.03 cutoff.
My actual take is that we should be ditching the BaC infraction entirely and just making "driving above .08" an element of reckless endangerment statutes like it should be from the beginning. Send these people to jail for a crime, don't write meaningless tickets.
I have a completely opposite attitude: driving should be a rare and exceptional privilege for the 10-20% of people who are best at it, and the rest of us should have to rely on professional drivers for transportation.
Drinking and driving should be treated the way pilots who drink alcohol are treated - not allowed to touch the stuff eight hours before takeoff.
But I know, I'm weird and the US would have to completely restructure society to make that work.
"the US would have to completely restructure society to make that work."
I think this undersells the challenge dramatically! Have you ever lived in a non urban environment? I'm not even sure how someone who lives in a rural or exurban environment would make this work without truly enormous subsidies.
Complete disagreement. Bright line rules have real value. The BAC level says to people if you drink multiple beverages and drive an automobile you are violating the law.
It is much easier for people to understand compliance with the law.
And .08 is extremely high and shows how much we write the law to placate drinkers and drivers.
One of the craziest articles I've read in recent years was an Atlantic piece arguing that it was bad that people were using Ring cameras to catch porch pirates. https://www.theatlantic.com/technology/archive/2019/11/stealing-amazon-packages-age-nextdoor/598156/ Notably, the article isn't claiming that there is some bad second order effect, they just thought it was unfair that people who were minding their own business, stealing packages were getting arrested for it. I thought it was nuts. There are lots of things wrong with the criminal justice system, but arresting people for actually stealing isn't one of them.
My wife still talks about how someone stole an iPod Touch out of her Amazon package 13 years ago, and has decided that that neighborhood was a bad neighborhood based upon it.
There's also the (extremely weak imo) argument that sometimes the people on the Nest cam videos will be misidentified. Which is true, but I can't imagine that happens _more_ often than without literal video.
And then there's this, which makes me want to puke:
> Despite the much higher cost of white-collar crime, it seems to cause less societal hand-wringing than what might be caught on a Ring camera, said W. David Ball, a professor at Santa Clara University School of Law. “Did people really feel that crime was ‘out of control’ after Theranos?” he said. “People lost hundreds of millions of dollars. You would have to break into every single car in San Francisco for the next ten years to amount to the amount stolen under Theranos.”
I suppose it's *theoretically* stealing when you invest in a fraudulent company, but if you really expect anyone to get as worked up about a person with enough money to invest in a startup losing said money vs a person whose kid's birthday present was stolen off the porch by a drug addict....your priorities are bit skewed to say the least. And in any other context except this twisted one where we are not supposed to punish racial minorities for petty crime would you suggest that the mom who just ordered a package from amazon is a less deserving victim than some investor-class white person.
Great article! I read somewhere once (can't remember where and a Google search didn't come up with anything) that a large percent of crime in America is committed by a tiny percent of the population. The subject of the story was a one-woman crime wave and excellent anecdotal evidence of that thesis.
Having been caught stealing does tend to make people more suspicious of you in regard to stealing, much as having been caught doing all sorts of other crimes tends to make one suspect.
It also tends to damage your credibility when claiming that you definitely didn't steal that unopened package, addressed to someone else, that you are carrying home, but rather found it laying on the side of the road and assumed it was abandoned property, free for the taking.
Sure- so long as it applies to LEOs too. My philosophy is that if you want to be licensed by the state to do violence on its behalf, you're also subject to a much higher bar for conduct. Felons have to give their DNA samples? No problem- also, every single law enforcement officer should as well. (It would have caught the Golden State Killer much earlier!) Parolees are surveilled more heavily? Sounds good- and also let's make sure every single cop in America wears a body camera (Mass state troopers still don't, last time I checked), that they have to be on *at all times*, and that the footage is immediately owned & stored by a completely different government body. No warrant required to start an investigation for bad conduct, by becoming a cop you agree that your conduct is filmed 24/7/365 on duty. Every cop's face should be photographed and added to a facial recognition DB, so that they can be instantly IDed when they commit crimes in public.
LEOs should be required to testify against each other as a condition of employment or having a pension. Do you know of a fellow officer committing a crime on the job? You should be *required* to testify against them, on pain of possible arrest and losing your pension. Perjury on official paperwork should be publicly prosecuted. All LEO internal docs (emails, etc.) should be immediately FOIA-subject.
I'm pro law & order! Let's apply law and order to an organized gang well known for carrying out brutal crimes and then concealing them, the government
There are one or two quibbles I might pick here (while agreeing with the basic thrust overwhelmingly) but I'm 100% on board with "I'm a law and order person, and that's why those charged with upholding law and order by the state need to maintain exemplary conduct!"
Think about all the money saved in the courts if we had video of surgeons screwing up, and reacted immediately rather than spend hundreds of thousands trying to obfuscate around it.
I'm sympathetic to the argument, but I don't know that it actually works that way. There's a lot of context around a surgery or medical treatment, and many judgements that often have to be made by experts. And oftentimes the judgements of experts will disagree with each other.
More transparency might not hurt in terms of exposing real malpractice and nefarious behavior. But it also exposes every move to Monday-morning quarterbacking from people who may not know what they are talking about.
Not just checkups - every estimate of medical malpractice I've seen comes up with hundreds of thousands of deaths per year. It's crazy to be so worried about police malpractice but blaise about medical malpractice.
I think it comes down to a central fact. People are much more comfortable with surveillance of those they consider "others." Rightists are fine with their tough on crime platform because it isn't their neighborhood that gets the oversight.
Left of center police critics are fine with this surveillance of cops because they would never be one nor do they think much of the people who are cops.
In both of your examples, I think more surveillance might be a good solution. But there's not a fair consideration given to the downsides to the people who would have to be surveyed. It's just like "F em", want to be a cop, deal with it".
You see this in attitudes toward teachers/police unions.
In both cases you have a coalition wanting to take away unionizing rights from a class of government workers.
In both cases they make good arguments that the services they provide are so important to society that unionizing is harmful. Police really can kill you. Teachers really can shape your life outcomes and your values.
But they only want to take unionizing rights away from the profession that belongs to the enemy tribe.
I think that we should just collect DNA from newborns -- I get that it has authoritarian overtones, but if you believe in liberal democracy and the rule of law, why not have a complete database going forward. It prevents all of the concerns about the bias in the criminal justice system meaning that some folks are more likely to have their DNA collected. There was a controversy in SF where a woman was prosecuted for some kind of property crime based on DNA from a rape kit she had provided in a previous investigation. Adding DNA collected in that way has obvious disincentives to reporting crime, but if we all knew that our DNA was in a database then we would all be on notice that it could be used against us in a criminal prosecution.
I really like the idea as a whole. I'd think a typical subpoena would handle requiring testimony against fellow cops. As for their own wrongdoing, are people allowed to give up their Fifth Amendment right against self-incrimination as a condition of employment? Even if not, forfeiture of a pension should do the trick.
I was in particular thinking of the shooting of Justine Damond in Milwaukee, where the shooting officer's fellow police refused to testify or say anything, and had to be subpoenaed by the prosecutor. They were- shocker- advised by the union not to cooperate with the investigation! That should be completely illegal- if you want to be authorized to do violence on behalf of the state, you should be required to 100% cooperate with every internal investigation.
'He said that his office started issuing subpoenas only after officers, acting on the union’s advice, turned down repeated requests to come in and provide statements.
“We tried again and again, and what they said was, ‘No, we’re not coming in.’ The Police Federation or its lawyers said, ‘No we shouldn’t,’ ” Freeman said'
Sounds good. Seeing as, as you note, the chief of the MPD was not willing to assist into an investigation of an officer shooting a completely innocent person, this just bolsters the common sense principal that a government agency should not be tasked with investigating itself. After all, if you heard about misconduct at the IRS or the NEA, would you trust them to investigate themselves?
So we can simply add this 'authority to order those officers to cooperate on pain of termination' to a civilian review board. And make sure that it's exempt from any union arbitration clause nonsense- if the officer declines to cooperate, they're fired immediately and no arbitrator can reinstate them, and they cannot be rehired for any LEO or government job anywhere in America. Plus, they lose 100% of pension accumulated to date
I am pretty confident that there's not one jurisdiction in the entire United States where a civilian run review board has the authority to compel to an LEO's testimony on pain of firing. (Not to mention that I explicitly included 'and this is exempt from arbitration', because I think we all know the government union arbitration game. Like 'America's worst cop', Florida LEO German Bosque- arrested 3 times, fired 7 times, he just keeps getting rehired thanks to arbitration!)
But I could certainly be wrong. I'm open to learning the name of even one municipality in America where a civilian review board can unilaterally fire LEOs. If you want to name even one....
From law enforcement to privacy to free speech and more, our policy mental models haven't caught up with or internalized the sea change in the facts of the world -- that we now live in a digital panopticon, where, like it or not, it's trivially easy to reconstruct what anyone was doing or saying, and where, from a collection of mostly privately owned but also public databases. Our mental models in these areas are still resisting accepting this fact, rather than accepting it as a given and the starting point that we should work with and use. For example, in a world of total surveillance and perfect enforcement, the more salient question about criminal laws become what they prohibit and what would be a reasonable (nearly certain) punishment. That's a different mentality than the "innocent until proven guilty"/"catch me if you can" crime deterrence model we've long had, a lottery model of very strict laws with draconian penalties, but poorly enforced.
I've wondered how traffic laws should work if everyone had a device on their car that recorded any time they were speeding. It's less obvious than it seems at first blush. We currently have a system that basically allows everyone to speed a little bit, but discourages you with reasonably large financial penalty if you get caught speeding more than reasonable. I don't think people would be OK with 100% enforcement, but I'm not sure what would replace it.
Speed limits conformed to road designs (and vice versa) could have a lot more legitimacy. When you build a highway and mark it at 30mph, then yeah, not okay with 100% enforcement.
I asked Matt a mailbox question about this that he unfortunately didn't answer. I'm wondering if car insurance companies will get there first by requiring transponders for their own bottom line desires. I'd also think you'd need universal transponders in order to make effective some other policies that are popular here, like congestion pricing, or replacing the gas tax with a per-mile tax.
At least Progressive already has a "Snap Shot" program where they use information from your car's computer about when and how you drive to set your rates.
I've also thought about how EZ Pass could serve this function. Particularly on highways where you pass through multiple gates, it seems like simple math - if the speed limit is 65 mph, and you go through two gates 65 miles apart in less than an hour, you must, at some point, have been speeding.
Traffic laws need a radical revamp. We wrote laws for speed limits and stop signs and the like that are simple to state, but don’t actually do what we want. So in practice, people exceed speed limits on most cases but stay under in others; they yield at four way stops without stopping; etc. The Idaho stop law is one of the first attempts to fix this, but only for cyclists. With real speed detection we would need to fix the speed limit laws too.
Disagree as to stop signs. IME they get a pretty good approximation of compliance much more than they don't. It's one of the high-trust / non-defection behaviors that America is actually pretty good about engaging in.
Depends on the stop sign. 4-way stops are notorious for low compliance and also confusion as to who yields to who. Ideally, you replace those with roundabouts, which naturally obviates these problems, but unfortunately in developed areas they are often too cost prohibitive to construct.
Sorry to necropost, but the British-style mini-roundabout (which is just a painted white circle) doesn't need any physical work to construct.
It functions essentially as a four-way yield with really clear rules on who you yield to (you yield to the road to the right, or the road to the left in the US where you drive on the right). Most drivers drive over the painted circle, you're not supposed to make a massive effort to avoid it.
Hm, I'll have to look into this one more. The challenge would be making those rules really clear in a place where they are unfamiliar--especially because typical 4-way stop rules are to yield to the car to your *right* in right-side driving places.
Replace 4-way stops with 4-way yields, so people don't have to come to a complete stop before moving off, and the approximation of compliance would become actual compliance.
This is what I *want* to say, but then I realize that the law doesn't have a concept of "4-way yield". A "yield" sign means that the one leg of the intersection always yields to other legs, and you can't have that at all four legs of a four-way intersection. We need three distinct signs for full two-way stop, for full two-way yield, and for four-way yield.
I'm not sure that you've sat and watched at a four-way stop. If you actually look at the wheels of a car at a four-way stop, you'll see that only a very small fraction of vehicles actually come to a complete stop for even a moment. The vast majority slow down, let the vehicle with right-of-way through, and then speed up.
This is exactly the sort of behavior you can do at red lights legally, if you time things right - slow down as you approach so that you don't arrive until it's your turn to go, and save some momentum so you don't waste gas. But even though we think of red lights as somehow stricter than stop signs, the official rules allow this behavior at traffic lights and don't allow it at stop signs.
Two way stops, where one street has permanent right-of-way, and the other always yields, do tend to have a lot more actual stopping, but I think it's still probably only about 70-80% at most, when traffic isn't high on the main road.
People obey stop signs as much as they obey speed limits - they do a good job of following the behavior that is not quite the legally posted behavior, but is what the rule should have been.
When I was a Boy Scout 60 years ago, I earned my safety merit badge standing two hours at a four-way stop recording how many cars came to a full stop. About 1% as I recall. Drivers generally have common sense. Which is why they also drive on highways at the speed of the traffic flow, even if it's 10 mph over the stated limit. They also generally obey the rule that at a four-way stop with two cars arriving at the same time, the driver on the right has right-of-way. I was recently at my wife's place in LeMans (France) where roundabouts have completely replaced four-way stops. The French know how to build them cheaply from the space in any intersection, and they greatly improve traffic flow. Bill DeBlasio, New York's unlamented ex-mayor, lowered NYC speed limits from 30 to 25 mph to reduce pedestrian deaths. No one abides by it.
I was recently driving in the south of France, and loved the roundabouts at every rural intersection! They also use them more at urban intersections than the US does, but after seeing it as universal at rural intersections, that clearly seems like the right way to go, rather than requiring a traffic light or a two way stop or a four-way stop, as we have all over the place in Texas!
Right, you'd want to relax speed limits on controlled access freeways (where the safest speed is often going the same as everyone else, and that tends to be about 10 MPH over what's posted), and make it stricter in places where there's a lot of non-motorist traffic.
Even better is to reconstruct streets to make it naturally prohibitive to break the posted speed limit, but that's a more expensive venture that takes more time.
I do a mix of agreeing and disagreeing about speed limits. Many drivers drive to the apparent speed encouraged by the road layout. In general, that should be how the speed limit is set, but there are cases where the speed limit is right and the road layout is giving the wrong messages to the driver.
There should probably be two sorts of limit signs, one that indicates "this is a normal 25mph or a normal 45mph, which is what we think you'd expect from the road" and another that says "look out, unusually low limit here; there's a reason for this, slow down; the road layout is lying to you". If the latter is rare enough that drivers will treat it as a hazard, then it would help a lot.
The second one should be, for instance, near elementary schools where small children are liable to wander out onto the street or where lots of kids arrive and leave at the start/end of the day, or where there is a blind bend just before a pedestrian crossing, or there's a hidden junction or whatever. But a combination of a hazard warning with a reduced limit would be a really useful sign.
There's a road near where I grew up which looked like it was an obvious 40/45mph road - straight, good visibility, no houses either side. But there was a bus stop one side and a pub about 50 yards in from the road the other side, and there were always people being hit when coming out of the pub to cross the road to the bus stop to get home. There isn't a hazard sign for "drunks like to cross the road here", but they dropped the limit to 30 just for that little stretch and installed cameras: far fewer people died. Lots of people who weren't locals hated this and complained because they thought the limit and camera were just to raise revenue.
Punishment is a wonderful thing. Not excessive punishment (definitionally bad) nor wrongful conviction (which isn’t punishment at all), but punishment fitting the crime, exacted by the state, after due process. It’s important for a basic moral reason: if someone has done wrong they should suffer the consequences. It’s also critical for a social reason. One of the key advances in the very existence of the state is the social contract in which individuals, families (and tribes/clans etc) surrender their claims to personal vendetta, even against the most heinous wrongs, and in return the state takes the responsibility to seek out and exact retribution on their behalf. This is hugely beneficial to society for a myriad of reasons. But we must not forget the hard basis of it all. The state has taken up the office of exact “vengeance” on our behalf. Thus punishment is an end in itself. Yes, we want to prevent recidivism. And yes, we want to deter others. But these are *additional* goals. They do not replace the basic point of punishment. So let’s not present punishment and enforcement as a zero sum game. In a well functioning society we have low crime rates, criminals are very likely to get caught, and then they are likely to get convicted and punished appropriately for the crimes they’ve committed. Its a sad state of affairs that MY feels the need to justify the idea of better enforcement per se, and that he presents laxer punishment as a carrot. Excessive punishment should never be tolerated, regardless of enforcement levels, nor should lax punishment.
"One of the key advances in the very existence of the state is the social contract in which individuals, families (and tribes/clans etc) surrender their claims to personal vendetta..."
Amerigo Bonasera (from The Godfather) nods in agreement
Your analysis doesn’t work for crimes that lack an individual victim. If Joe six pack drives drunk but isn’t that drunk and gets home safely, there is no vengeance to exact because no one got hurt.
Arguably there is - he put the safety of the public at risk. However I conceded in my original comment that this isn’t the *sole* reason for punishment, and it perhaps doesn’t apply in every case (that’s why we have additional, independent justifications, such as deterrence). However it is a foundational reason for the existence of criminal justice generally and must feature in every discussion related to the topic (or at least loom large in the background). I feel like MY’s post, like so many these days, are forgetting the basics.
I guess what I’m pushing back against is what I would call the dark side of puritanism— taking pleasure in punishing rule violations. This is a real human tendency, it is often called righteousness.
The state should be more like the temperate parent reluctantly spanking its children than the sadist who takes erotic pleasure in spanking.
“Punishment is a positive good in itself” is one step from getting off on spanking evildoers.
I think the parent analogy is totally false. The spanking is for the benefit of the child. Punishing criminals is not for their benefit but for that of society and the victims. I think it’s very healthy and human to take a measure of satisfaction or relief and yes, in certain cases even joy, in seeing Justice served: that Justice be served , and be *seen* served is crucial for society’s well being. However this response is of the victim and the people who care about them, and the public at large. The *application* of criminal Justice should not be emotion-driven and that’s why we have laws, professional judges and lawyers etc.
The problem with Puritanism as you put it is in over criminalizing things that ought not be criminalized and in being religious fanatics. a totally separate issue.
part of me agrees with you. this part also thinks public executions would be really cool if the condemned really were degenerate murderers.
a few well chosen executions per month could slack the public’s need for vengeance admirably. there could even me a democratic, participatory process to decide which death row inmates are actually executed.
i’m not sure this would work out well, but i’d be happy to try. it’s certainly a more elegant way of pacifying the demos than mass incarceration. i’d rather have public executions of murderers than long sentences for every 19 year old burglar
"a few well chosen executions per month could slack the public’s need for vengeance"
I don't think it works this way. I think if you whet the public's appetite with some public violence, they'll happily demand more to feed that appetite.
I agree. And one of the most immediate things one could do to implement it is move traffic enforcement and public nuisance events to non "police" officers (and to technology, including enlisting citizens to record parking infractions) so as to let "police" officers focus on crime deterrence. apprehension of suspects, and provision of evidence for prosecution.
Sending someone a ticket in the mail is not enforcement unless they actually pay the ticket. At some point, you need to be able to boot cars, garnish wages, arrest people, or otherwise create actual consequences, which hits up against the same problem of "some people are actually against enforcement."
But most people DO respond to mail fines so that frees resources for booting, garnishing, etc. In DC we have the additional problem that VA and MD do not honor our tickets (although I'd think that reports to credit bureaus of unpaid fines would be somewhat effective).
"But DPW only has a staff of four right now operating in two booting teams, according to acting DPW Director Christine Davis. They can only boot about 50 vehicles a day, meaning it would take more than 25 years to get through a backlog at the current rate."
So we're only doing the sending tickets part -- in theory we could be freeing up resources, but that doesn't seem to be how this is playing out. Certainly MD and VA are a huge part of the problem, but it's not just them, and the fact that DC outlawed suspending licenses for unpaid tickets is also a part of the problem.
So each of the four booters is booting 12 a day. This sounds like a management/labor problem (lazy booters) or a technical problem (invent a better boot). I assume more booting will solve the MD/Va jurisdictional squabbles.
In the context of the vast amount of lack of enforcement of other traffic- and non-traffic related issues, I would conclude that while the mechanism for lack of enforcement may be something like "labor problem", the cause of the issue is "the city doesn't want to do enforcement."
Not suspending licenses for unpaid tickets is probably a mistake, but suspending licenses does not automatically get fines paid, either. Sounds like more booting would be a good use of the resources freed by more use of technology. And/or just an increase in resources; it ought to more than pay for itself. How about contracting with private towing companies?
There are all kinds of policies we can probably agree on! My point is that the same forces which drove lack of police enforcement are also driving lack of other enforcement, which is that the government does not want to be doing enforcement. This isn't a "if we just did this in a way other than cops" issue. We had police doing enforcement, and now we have nothing doing enforcement, unless you are the sort of person who pays your traffic tickets even if nothing is making you pay your traffic tickets. So the idea that if we do less police enforcement that will free up resources, or there's some technocratic fix here, while true in some sense is not true in the sense of actually being what happens.
Yes and no. Sure, no policy we might want to change just popped up with no politics behind it. But the whole point of Matt's essay is to try to change the politics, that, for example, we got specific non-enforcement bias from a misunderstanding of how best to reduce incarceration. If we think every issue is in a perfectly locked socio-political equilibrium, we are all wasting our time writing about it. :)
Just to hijack (carjack? :)) the conversation, technology exists to document and fine excessive muffler noise. It should be used but those non-"police," traffic officers. And although fines are not appropriate, excessive serene noise should also be reported to police, fire, and ambulance company headquarters.
It may not be oxymoronic in the future: when electric cars become more ubiquitous, there's a real concern of them being too quiet for non-motorists to detect and safely avoid.
That really is an easy problem to fix. You just put a piece of carboard on the frame so it makes noise as the spoked wheels turn, like we used to do with our bicycles. :)
Yes, do fine them. And I'm not opposed to "police" officers stopping cars with (maybe more than one) outstanding violations.
I'm rejecting the idea that using technology to detect and send summands for traffic violation is useless. that only armed police officers can issue meaningful summons. Really, what's the difference between an officer-issued summands and a machine issued one? The person still has to pay by mail or show up in court?
Toll roads don't use officers in cars to issue tickets for not paying tolls.
He’s talking about sirens there. Maybe confiscating the sirens from the ambulance of a private hospital that is abusing them is appropriate, but it’s tantamount to putting them out of the emergency business.
Not necessarily. Sirens are OK. Not all are excessive. But they do not need to be audible3 and 4 block away for traffic to accommodate the emergency vehicle.
**PUBLIC** surveillance cameras, i.e. covering public spaces, are a good thing. The more of them the better. Nobody has a reasonable expectation of privacy in public.
Another way to think about it: if it would be OK for a cop to stand and look in the same direction as the camera, there's no privacy objection.
I can stand on the sidewalk and watch who comes in and out of your house, but a) you can see that I'm doing that and b) it's labor-intensive. If we were to move toward tech that could observe that cheaply and ubiquitously for everyone, I think that would have some dramatic implications, and I do not think that fears would (or should) be calmed just by saying "public space."
As I've said repeatedly in these comments the fundamental problem here is that we are over-criminalized and under-enforced
That is to say that our criminal code includes vast amounts of nannystate, social engineering horseshit like vice taxes and pretextual offenses and victimless crimes and municipal revenue generators that are not rightly criminal that create toxic incentives and vast opportunity for oppressive, discriminatory, corrupt and ineffectual enforcement.
We are also obviously under-enforced in that we don't reliably catch murderers and car thieves or intervene when some addict is ransacking the CVS, largely because policing primarily does those other things.
The problem with Matt's vast surveillance state is that he's regularly shown that he thinks the technocratic, nannystate, social engineering stuff is good actually. As long as people are committed to using the criminal code to manipulate the law abiding, instead of to reliably prevent and punish crime this sort of proposal should be opposed, tooth and nail, by anyone who values a free, pluralistic society. The threat posed by Matt's preference here is vast and while I agree that it offers effective enforcement opportunities, so long as the guardrails on WHAT these powers might be called upon to enforce are so obviously compromised it cannot be allowed.
A criminologist friend of mine told me years ago that it is the odds of getting caught that had a demonstrated deterrent effect, not the consequences of getting caught. He was right.
Like most, I’m all in favor of catching more criminals but as things currently stand, that may not do much. In many areas, those caught for property crimes and even some assault and battery charges are out the same day they are caught (or the next). Prosecutors and police have given lower priority to such crimes due to various reasons, some political (requiring bail exacerbates the disproportionate racial effect) but also because of the sheer volume of such cases. There simply aren’t enough jail cells, guards, prosecutors or judges to increase the certainty of prosecution and conviction. Using all the technology that might lead to more arrests won’t necessarily result in increasing the deterrent effect of prosecution, conviction and incarceration. Unfortunately, bringing down crime rates is going to be much more difficult than taking DNA samples and enhancing surveillance.
Assuming criminal behavior is rationally motivated, increased apprehension ought to reduce crime by reducing the net benefit of the activity. However a substantial portion of crime is non-rational as a consequence of substance abuse among other things so cost-benefit calculations are irrelevant.
The bigger issue for me is that a pervasive surveillance society tied to facial recognition is subject to horrifying abuse. C.f. Xi's China. It is also true the the UK has had a pervasive CCT network without sliding into fascism. But the potential for abuse is clear. The only contemporary Republican whom I might trust with such a system is Mit Romney. There are also many Democratic officials who couldn't be trusted with it.
Fully rational crime would be deterred by the product of harshness of penalty and probability of getting caught. Purely irrational crime can only be deterred by understanding the weird irrational way a being operates. Most crime is partially rational - people are deterred by probability of getting caught, but aren’t particularly deterred differently by a one year sentence vs a thirty year sentence, or a thousand dollar fine vs a million dollar fine. Increasing probability of getting caught while proportionally decreasing punishment does generally increase the deterrent effect, though not to fully rational levels.
The way I've seen this expressed elsewhere indicates that, interestingly, criminals as a class are somewhat different in kind from most people: most humans are risk-averse but there's evidence that criminals are disproportionately risk-preferring.
I just think you're confusing the tools with the will-to-use-the-tools. If the US wants to imprison people for political activism or being an ethnic minority it can already do that. Likewise, China was locking up dissidents and Uyghurs long before they got ahold of facial recognition.
If the only thing Governments and police forces did was oppress people then I would support limiting their technological prowess. But the police also arrest people for crimes, and there are tens of thousands of fugitives out there, wanted for everything from rape to murder to kidnapping to just jumping bail. It would help our society to decrease that number.
And with an authoritarian movement gaining strength in this country, the will-to use-the-tools can be assumed. We are on the edge of a precipice. Expanding government surveillance powers at this time is inadvisable. I thought this was clearly implied in my original comment.
That's not what most people mean when they call Trump authoritarian. MAGA people want to do a lot of things, but setting up a dissident-squashing repressive surveillance state does not seem to be one of those things.
I suspect many of the opponents of the use of facial recognition technology (but perhaps not all of them) are not opposed to the enforcement of criminal codes,but rather very skeptical that governments can be trusted with such tools. Considering the already highly intrusive surveillance state apparatus that exists in the five eyes countries - the thought of adding in facial recognition technology is quit terrifying.
With cameras everywhere this could give you a capacity to track where everyone is at any moment - especially if everyone is installing doorbell cameras etc whose corporate suppliers are likely to bend to rubber stamped warrants or make deals with the NSA to further government surveillance.
There's some of that, but there is also an "enforcing the law is bad" crowd. There was an Atlantic article a few years ago that was aghast at the fact that Ring as were resulting in people being arrested for stealing packages.
https://www.theatlantic.com/technology/archive/2019/11/stealing-amazon-packages-age-nextdoor/598156/
That was a long and wild read. I did have a caveat in my original comment, in anticipation that some people might actually hold the view that MY discussed. I have certainly heard about progressive jurisdictions where theft just isn't prosecuted, but the whole concept seems so crazy it's still a bit hard to believe.
Yeah, it's nuts. I understand that there are reasons why police might not go all "Inspector Javert" over a couple stolen packages, and that we may not even want them to do so, but the thesis of this article is very much that it is wrong and, perhaps racist, for people to get upset that someone is repeatedly stealing their Amazon packages.
It was certainly a unique experience to witness a writer try to spin the 27th time someone’s engaging in theft.
I suspect that a regime that is going to illegally abuse facial recognition is also not going to be stopped by a law against using facial recognition at all. We might as well have it as a tool for legitimate law enforcement investigation because I doubt that some "ban" on it as a technology is really going to save us from Big Brother.
If we ban face recognition, then only criminals will recognize faces.
Right. The slippery slope argument is bogus. If there's any autocracy in America's future, that regime is going to use whatever tools technology makes available. And there won't be any debate.
Those arguing against greater use of these identifying technologies need to point out why rule-of-law constrained governments and police shouldn't use such technologies. There are arguments that can be marshalled, for sure. But "it puts us on the path to Big Brother" doesn't seem like one. What puts us on the path to Big Brother is things like elections nullification, and subversion of democratic norms.
Keeping it illegal does afford greater protections for citizens than not - for example you could claim fruit of the poisonous tree, to get evidence excluded where it was derived from the illegal use of facial recognition mass surveillance. Similarly, it limits options for funding and deploying the technology.
Couldn't one legislate what face recognition can be used for and what not?
Then it is clear in advance what is legitimate use and what would be fruit of the poisonous tree.
(thanks, UK, you made me look this up and I wrongly thought it was a "poisoned tree")
Also: if we rely on face recognition, we should make sure that the judicial system is able to find exculpatory evidence on this basis: if you're accused of a certain crime at a certain time and place, your defensive team should be able to look for evidence of your presence elsewhere (which is not easy as recordings are often deleted for reasons of privacy and storage capacity).
You are perhaps over-estimating how expensive this software is, or at least how expensive it will be in the not too distant future. More than likely, if an evil repressive government wants this technology it's not going to be hard to get it.
Agreed. Let’s not kid ourselves that the database wouldn’t also be full of Trump supporters merely attending a rally or abortion activists marching in protest. This information could easily be used to harass opponents of the ruling political party, justifying searches and detainment just because having a match in the database provides cause.
Either you have a society where the government is empowered to harrass people for their political views or you do not. The Soviet Union was quite the archetypal totalitarian state without a lot of high tech tools. I don’t think today’s China has anything on 1950’s Russia in that regard.
>>Agreed. Let’s not kid ourselves that the database wouldn’t also be full of Trump supporters merely attending a rally or abortion activists marching in protest.<<
I don't think we need to "kid ourselves" this wouldn't be the case. We need to pass laws and formulate regulations preventing this from being the case. If we can successfully do this, why not opt for less crime? Are you claiming we cannot do this? Perhaps access to the database would be provided only upon issuance of a warrant...
I am claiming we can not do this.
https://en.m.wikipedia.org/wiki/Room_641A
There are things that people want to do that are ethical but they don't want the whole world to know about it.
Maybe they are into some consensual but potentially not-well-accepted sexual interests. Maybe they go to AA meetings or a psychologist. Maybe they are involved in a controversial political group. Maybe they are in an insular religious community but are considering leaving. I can think of millions of reasons why someone wouldn't want all of their daily whereabouts monitored.
We don't want to live in a world where the government has so much surveillance information that they can easily harass, intimidate, or blackmail someone that they don't like with this information. So if we are going to increase survaillance for better enforcement of criminal laws, we need assurances that these measures will be only used for investigating crimes, and that authority to access the data is highly limited both in terms of who has access and how the data can be used.
Who is going to look at the results of this data from every camera on every corner in every city and town in the US?
There should be regulations that confine analyzing this data, that doing so without a clear reason is unacceptable (cause for dismissal.) That's the case for hospital employees; you cannot look up a patient's information without a good reason.
My reading was that Matt was advocating for *more* cameras, which I assumed would be government supported. I don't imagine people would watch this footage for casual fun; they would be looking for a criminal. Therefore anyone who does not seem to be a criminal would not be subject to facial recognition technology, which I guess is the basis for people's objections. Not sure if they're worried about being put in some kind of national facial recognition database for future punishment/ oppression, or about giving criminals an unfair disadvantage?
I do think that there's a difference between a world where there are cameras in some public places, with the data only being able to be parsed by a human is different from a world where cameras in public are nearly everywhere, and your face can be automatically recognized and all the data can be automatically processed, analyzed, integrated and stored at a large scale. Perhaps the latter world requires some reanalysis of the pre-existing social contract
Yet the dream of most Twitterers is to have a post “go viral” so everyone sees it…
And just to add it shows society does a pretty good job ***not*** sharing this info.
You almost never see videos of regular joes blowing their life saving in slot machines for 8 hour shifts even though the videos exist. Nor do you see videos of guys often slurring their words. Occassionally some pure schmuck gets exposed. But by and large we never see this stuff.
This is a good point, but the kicker is less about the government having technology like this and more about the fact that the technology itself is extremely opaque and cops, attorneys, and juries are all going to be equally unable to engage with it productively. Law enforcement is just not equipped to deal with this.
How do we monitor the algorithm? Who gets to pressure test it and make sure that it is doing what it says it's doing? What happens when version 3.4.96 identifies your kid in the grainy screencap of someone getting out of a car, and version 3.4.97 disagrees?
I strongly suspect that juries will just accept "computer says it was you" as an extremely compelling argument, because that's how most people interact with technology.
I work in ML/AI and volunteer with a group that encourages BIPOC young adults to learn about and pursue careers in math/science/technology (I mentor in ML/AI/scientific computing). What you're saying is exactly what someone in the know should be worried about with facial recognition. It's not that good, it's lazily opaque to the average citizen, and it's accuracy diverges significantly by the race, gender, and age of the person being classified. Guess which groups it performs the worst for -> https://sitn.hms.harvard.edu/flash/2020/racial-discrimination-in-face-recognition-technology/
I don't think the main point of the technology is to decide who is who on a video once it gets to court. I think the idea is more "a car was stolen yesterday, we have video showing random dude doing it, we have no idea who he is, lets run the picture through a database to see if a name pops up" at which point you use that as a lead to find them and then use the video itself as evidence and let the jury decide how much the accused looks like the person on the tape. It's like a faster and more certain version of releasing the video to the media in hopes some viewer recognizes them.
Yeah, I think we might disagree here. I think that the scenario you describe sounds plausible, but note the basic circularity of the logic that you are using:
There is a video of someone that committed a crime, and the cops search a large database to find a person that looks like the person in the video. The database is large, so the computer works hard and finds someone (or probably a lot of people) who plausibly look like the person in the video. Then the cops present the video as evidence that the person that they arrested committed the crime, and asks the jury to decide for themselves whether the person on the stand looks like the person in the video. And he probably does, because the computer is pretty good at finding people that look like that. But did the computer pick the right person among all of the people who plausibly look like the video? Is the right person even in the database to begin with? There's not really any way of knowing without secondary evidence.
In a lot of cases the cops might find a lot of secondary evidence when they investigate whomever the computer selects, which is fine. But in a lot of cases they won't, and in cases like that I'll bet there would be a real temptation to use the exact same logic that you did above to get a conviction anyway.
Fundamentally though, how is that a more likely result than a similar misidentification from broadcasting the video on the local news and having someone call in and say "that looks like my co-worker!" The technology is really just another tool to find the suspect and at the end of the day the plausibility of that being the person in question has the same burden of proof as it would using contemporary methods.
Again we might disagree, but I think it's basically because in practice people give a huge amount of deference to technical solutions because the underlying issues are too abstract to reason with.
If a person gives a tip to the cops the defendant's attorney can cross examine the informant and ask a bunch of pertinent questions to clarify their basis for identifying the person in the picture, which amounts to additional evidence supporting or disputing the ID. None of that secondary evaluation of the ID itself is really available in a database search, because an attorney can't ask the convolutional neural network why particular layers were tuned the way that they were.
So put up guardrails and trust that the USA is fundamentally different than China.
The biggest factor to unsolved crime is the "street code". If there's a crime and no one in the neighborhood wants to talk, put up cameras as a last resort. This kind of thing could go a long way to reducing inner city violence.
So which guardrails...isn't that the question?
And I don't make that assumption--not by trust. We are only different than China if we continually work to make it so.
If the govt abuses the technology then vote em out.
This is error correction and why liberal democracy can be trusted to experiment with technology while dictatorships cannot.
No, rights, like privacy, should not be up for majoritarian rule.
A law that only allows previous felons to be tracked.
Many such guardrail possibilities.
Laws that prevent broad-based facial tracking. Only allow facial recognition to be employed when there's a crime suspected.
This is an intuition I hold, too. Laws on the books we're not going to enforce are actively harmful. The Anker paper backs this intuition up from another direction.
This is my hang-up. I agree with most of what Matt's said above, but giving the government the tools to track its citinzenry's every move... Let's just say there's a reason China is the main exporter of this technology/civil model.
I broadly agree , but I think this piece’s main weakness is that it’s written against the backdrop of Twitter crazies who oppose law enforcement. Thus it doesn’t address the concerns of normal people who instead prioritize the well being of the average, law abiding citizen. Us normies are all for catching (and punishing !) criminals , but we have other concerns about expanding police powers. How do you make sure innocent people are not falsely accused? How do you balance these real needs with privacy concerns (e.g. do we allow any and every cop unmonitored access to the most powerful facial recognition tech?). In short, in a healthy democracy we should always take pause before giving the government new powers. Ultimately, as often, we may conclude with a sigh that giving this power is a necessary evil, but we may want to think about adding some checks for reassurance. I wish MY would have addressed these mainstream concerns.
For exactly the same reason, we should ban the use of eyewitness testimony. It’s the source of a huge number of false accusations every year.
I can't speak for every jurisdiction, but in the ones I am familiar with, courts absolutely are aware of the low reliability of witness identifications, and the rules of evidence have been re-written to address and mitigate these concerns.
I hadn’t heard that, but if it’s right it’s going to lead to even fewer criminals being punished unless we can make up the gap with more reliable tools.
eye witness testimony requires less lawyerly skill to introduce at trial than any other form of evidence. rules against hearsay make it difficult to testify to what a witness learned from others. what they personally saw is fair game unless it is completely unconnected to the charged offense.
one can certainly call an expert witness to impeach identification testimony, but few criminal defendants can afford experts. i have one client who has a very good job and she is struggling to pay for a single expert. she’s also one of my better off clients.
This might be the case where you live but is not reflective of other jurisdictions.
georgia has adopted the federal rules of evidence with only token amendments. we are quite representative
Can you vaguely indicate which jurisdictions these are? I'm an NA lawyer and while we're vaguely aware of the sketch nature of eyewitness evidence at least my jurisdiction still heavily favours it over other forms of evidence.
No, but we do have strict rules about evidence admissibility, perjury laws, the critical role of cross examination etc. centuries and millennia of experience taught us how to use witness testimony correctly and beware its pitfalls. We should aim for no less when incorporating new methods.
IMO the Federal Rules of Evidence and their state analogues are not by any means paragons of the correct application of evidence (inter alia they're at least in part explicitly counter-Bayesian) and reflect the staunch reliance on tradition to substitute for considered and principled reasoning. There's a prominent, and in my view unfortunate, strain of thought that essentially treats Anglo-American evidentiary norms as good by dint of their persistence and pedigree rather than because they bear the slightest relation to how actual humans conduct themselves and make assessments outside the confines of a court of law.
Not everything in them is bad, mind you, but treating them as the considered result of decades and centuries of evolutionary pressure to reach the correct result is a long way from the correct paradigm in my view.
omg yes. on the other hand, the first thing clients want to do is dig up/drag in dirt on their accusers. the instinct is very strong and transcends race, gender and class. blah blah blah uses drugs, had an affair, whatever, so predominates what i hear it can be hard to focus clients on their actual cases.
also, the legal definition of relevance (rule 401 i think) is very very broad. the principle that probative value just outweigh potential for unfair prejudice is 1) obvious and 2) too i’ll defined to be much more than a blank check made out to trial judges
In the one case where I was selected for the jury, I was shocked to see that the police officer who testified was not allowed to read from his notes while testifying, but was allowed to look at his notes to "refresh his memory" and then testify again after doing that. I can see that there's some intuitive idea of how the memory works on which this set of restrictions and permissions might make sense, but not on any realistic idea of how it works!
The one jury I've been on, two cops testified in identical words. I was elected jury foreman, and we acquitted in five minutes.
Can you say more about the counter-Bayesian part? I feel like I almost get it but want to be sure I’m understanding you right.
If you have JSTOR access, I think this paper by Laurence Tribe does a remarkably good job of arguing that there are certain ways the standards of evidence *should* be anti-Bayesian: https://www.jstor.org/stable/1339610
And I say this as quite a committed Bayesian!
The biggest and most prominent examples are FRE 404(a)(1) (generally referred to as "conduct in conformity" [with the defendant's character]) and FRE 404(b)(1) (disallowing prior bad acts as evidence of a defendant's character). https://www.law.cornell.edu/rules/fre/rule_404
(a) Character Evidence.
(1) Prohibited Uses. Evidence of a person’s character or character trait is not admissible to prove that on a particular occasion the person acted in accordance with the character or trait.
[...]
(b) Other Crimes, Wrongs, or Acts.
(1) Prohibited Uses. Evidence of any other crime, wrong, or act is not admissible to prove a person’s character in order to show that on a particular occasion the person acted in accordance with the character.
FRE 404(b)(1) is such a shitshow of a rule that it was eventually abrogated in part as to sexual offenses (FRE 413, https://www.law.cornell.edu/rules/fre/rule_413)
Separately, the hearsay rules more or less in their entirety. In brief: the world you and I know and live in and all of private enterprise basically run on hearsay (modulo the fact that technically hearsay only exists outside court of law and thus literally everything outside a court is hearsay), and in courts of law it's presumptively disallowed.
I think this is all kind of addressed, at least implicitly. The thesis is that more effective policing means less crime means fewer accusations at all, which should lead to fewer false accusations. Better technology use should lead to fewer false accusations and/or easier exonerations (as we’ve seen with the expanded use of DNA evidence).
I get the apprehension about giving the police more tools, but I think you need to make a case of what they’re going to use those tools for that’s different and worse from what they already do. I think that’s an easy to make when the new tools are military weapons and vehicles. With surveillance stuff? I’m not sure how much would change given the degree to which we’re already tracked by cameras and internet stuff. What would they use it for that they can’t do already with eyewitnesses, patchwork video availability, etc; and why would it be bad?
IMO it’s addressed only under the implicit premise of a police force that’s 100% professional and in good faith. I’m far from those demonizing the police, but let’s not forget that they’re human, that power corrupts etc and frankly that beyond the theoretical concern evidence isn’t lacking for abuse of current police powers by bad actors on all levels. I think these concern deserved explicit and comprehensive attention in the piece.
But what part of this involves giving the police more power or relies on them being 100% professional and in good faith? They can already arrest whoever they want with lower standards of evidence. They can stop and kill people with relative impunity. The people that they do investigate, charge, and convict end up in prison for a very long time! IMO one of the underlying assumptions here is that the current police aren't actually that good in the US so we need to find a way to make them better. So let them use facial recognition instead of relying on eye-witnesses and racial profiling. Use ankle bracelets and breathalyzers so there's less differential ("racially problematic") enforcement of bail/probation conditions. Speed cameras instead of DWB! (I know that last one is from a different piece than here, but it's in the same spirit).
Assume that the police are... sub-optimal (I certainly do!). But then tell me how these proposed powers (and I'm not sure how much of this I would classify as new powers per se, but ok) would lead to them doing worse things. As I mentioned above, this is easy to do for the military stuff: If you give cops more guns they will shoot more people. If you give them tanks and shit they will act more like an occupying force with any provocation (including riots or whatever, but also even just peaceful protests). Even if you just give them cool tactical gear (helmets, vests, cool belts, etc), it will encourage the cosplay warrior crap which makes them less effective and more dangerous to the public.
Can you make a similar case against using facial recognition and location monitoring? I know that sometimes people use "don't break the law if you don't want to be punished" in bad faith. But I think the point here is that if you have effective crime detection, then you won't get falsely accused of breaking into houses if the GPS shows that you weren't breaking into houses. Or if you don't drink above a certain amount then the cops won't waste time following and harassing you to check up on whether you're complying with your probation/parole conditions. So what is the case for "using these technologies would let the cops do new bad things" instead of doing fewer of the old bad things?
To be fair, if you're just of the opinion that they'll find a way to abuse anything given the chance, I 100% won't argue with that. But I do think that there are levels of degree, and that what's proposed might lead to less police misbehavior than the status quo.
If police officers are too "fucking stupid" to use valuable tools to reduce the crime rate the solution is to recruit smarter people into policing not ban the use of valuable tools.
This doesn't make sense to me as an argument against more non human surveillance. If you don't trust the police, wouldn't you want there to be evidence outside of a detective saying "he did it?" If you a juror, would you feel more confident convicting based on eye witness testimony or on camera/DNA evidence?
Neither, or either . I’d be confident convicting based on the weight of the accumulation of the evidence , the more the better, but mostly the better scrutinized the better. Camera/dna evidence isn’t some objective truth immune from heuristic challenges and manipulation and it requires rigorous methodologies of application just like witness testimony.
In any case all this is a separate issue from the question of abuse. The answer to that is simple. It’s not black or white. You don’t write someone a blank check but you do trust them with a signed check for a specific sum despite that also compromising you to an extent (your routing and account numbers are on there etc). Life requires balance. With government we need to try to get good people there but also put guardrails on them to guide them to do their work properly and not abuse their powers.
I suspect that a good deal of the concern about facial recognition software is rooted in a distrust of the police and how they will abuse it.
Other than senior citizens whose opinions are shaped by television police procedurals, a lot of people (and not just Twitter crazies) believe that police are prone to abusing their authority -- particularly when it comes to interactions with Black people.
Because of that distrust there is probably a stronger than is actually warranted reluctance to give police even more tools that could be abused.
I think the trust/distrust dichotomy isn’t helpful. There are very few angels or devils out there. We need to rather think in terms of cost benefit analysis (as MY does) but then *also* account for potential abuses both within and outside the legal procedure and regulate to minimize them (the issue MY overlooks).
P.S. I strongly recommend reading John McWhorter on the police and black people. Some of the worse things most liberals now take for granted are simply factual untrue. It’s a very sad state of affairs.
"this piece’s main weakness is that it’s written against the backdrop of Twitter crazies who oppose law enforcement."
I don't think the idea remains with just "Twitter crazies." In my undergrad, that ethos infused many anthropology/sociology/ethnology/South Asian studies talks, because the professors in those disciplines regularly worked in non-democratic or authoritarian countries for which the legal system did not accurately reflect the practices of the populace, and greater enforcement means greater hardship. I suspect the principle then diffuses out to people who study the US legal system, without anybody really paying attention to its preconditions for relevance (also relevant: Matt's discussion of trans-Atlantic pollination in last week's mailbag).
But yeah, in a democratic society, the concept doesn't really make sense.
I never meant to imply that Twitter crazies don’t have day jobs, but that’s not how they’re influencing MY and skewing the emphases in his public writing.
I think we need to talk about the elephant in the room: if catching criminals is good, it will involve catching a disproportionate number of black men.
Progressives don't like enforcing crime because it will target black people much more commonly. If we use facial recognition software, or make more extensive use of DNA databases, they will be used much more commonly against black men. Of course, if these tools ultimately reduce mass incarceration while reducing crime, then everyone is better off, including the black men who would ultimately end up going to prison. But progressives will fight these measure tooth and nail. Expect articles in Vox about racially biased facial recognition, for example.
Liberals would be better off talking about this stuff, just as conservatives would be better off talking about a coherent health care policy.
There’s no elephant, and this fact was always well known. The other side of the coin is that black communities will disproportionately benefit because they are disproportionately victimized by crime. Such policy would thus be neither racist nor “affirmative action” but simply the state doing its job to protect its citizens. The problem with disproportionate black crime has root causes that need to be addressed as such but the idea that giving black criminals a pass on racial grounds is a legitimate or helpful measure let alone some “anti racist” necessity is pure insanity that thus never occurred to anyone before the last decade, and it’s still totally unclear to me how this toxic nonsense ever came to mainstream attention.
I think you underestimate the elephant content of the room. I've encountered progressives in real life, not just on Twitter, who are genuinely willing to take the position that wealthy white people commit *violent felonies* at least at the same rate, or possibly even higher rates(!), than poor black people, and that the reason that's completely unreflected in crime statistics is that police and/or prosecutors cover up this wave of white criminality, if it ever gets reported in the first place (which they believe in many cases it isn't).
This is especially popular among the conspiracy theorist left.
Someone made this argument to me on Reddit the other day! They added the additional theory that rich white neighborhoods were intentionally under-policed for this exact reason.
Yep, I've encountered that variation in real life too. It typically comes up when someone is explaining how it's racist that there's so much more police activity in [POOR NEIGHBORHOOD] than [RICH NEIGHBORHOOD] and I respond by pointing out that [POOR NEIGHBORHOOD] has dramatically higher violent crime rates, to which the other person replies that [RICH NEIGHBORHOOD] is really just as violent, but the police stay away from it because it's full of rich white people. If I proceed further and question why rich white people would tolerate all that crime in their neighborhood, I usually get a response that sounds like some variation of "Get Out"/"The Purge" where the rich white people are supposedly beating/raping/torturing/murdering non-white workers or random passers-through.
My experience of the "black community" indicates that today it's still very much the same community which backed Clinton's '94 crime bill to the hilt.
The hipster lefties just want to pretend that the 12 black humanities Ph.D grads that they follow on Twitter are "representative" of African Americans, of whom under a quarter have a bachelor's or graduate degree, and have erected a circular firing squad and peer pressure architecture around that stupid shibboleth.
I've always thought that there's a niche waiting to be filled in The Discourse for a Black commentator pushing back against those 12 Black humanities Ph.D grads. There's probably already someone out there that I'm just not aware of. Someone like Ruben Gallego telling people to cut it out with the Latinx crap.
John McWhorter
Roland Fryer, Thomas Chatterton Williams, Kmele Foster, Glenn Loury, Coleman Hughes, even Adolph Reed. There’s no shortage, but our gatekeepers on the liberal side seem to have decided that they’re all suspect somehow.
Hence the quotation marks, yep.
No ethnic or racial grouping within the US is a monolith.
But I regard it as a particularly stupid exercise for the woke/hipster left to anoint a very small, skewed sample as representative of such a "community."
“They are disproportionately victimized by crime” really is something that we don’t seem to talk about like we should, and I find it heartbreaking. The vast majority of people in zip codes where bullets fly frequently are law abiding citizens. Think about the way we talk about school shootings, and think about the way we *don’t* talk about being a person trying to raise a family in those neighborhoods. It’s horrifying. School shootings terrify us and dominate conversation even though they’re statistically rare. There are places in this country where you really do need to worry about shootings, not a statistical rarity at all, and we (those of us who don’t and won’t live in those places) just sort of take it for granted.
The people in that situation have less representation in our national discussion about crime than the criminals do. If white people with money had to live that way, this country would already be fascist because if it took a police state to stop it, we’d have a police state.
It’s like we’re so afraid that some conservative will bring up “black on black crime” that we make little or no effort at helping the people who are being terrorized. Again: the vast majority of people who live in our nation’s scariest zip codes are 1) black and 2) law abiding citizens. How is this not one of the most pressing concerns of people who care about equity?
Because they don’t actually care about black lives but about virtue signaling and policing their fellow liberal whites?
Also, given that perpetrators of black on black crime are less likely to be caught than perpetrators of black on white crime, and given that most murders are intra racial, more enforcement will mean more black people going to prison
If punishment is like lightning striking, if the randomness of punishment blunts deterrence, and if a lot of black people are being struck by lightning, a sensible anti-racist might want lightning strikes to be gentler.
What, you want black-on-black crime to be punished less severely?
Do you not value Black Lives?
(inevitable insane-prog follow-up)
You are perhaps facetious but I would make that exact response , word for word (sans the parenthetical) in earnest.
I literally made that exact response on the traffic camera article a while back.
Yeah. It's a facetious/bad-faith interpretation of what was said above.
But it actually would be a legit response to the suggestion in some ways.
It would also just be a 180 from their standard response, but I could seem them being goaded into it, regardless of the I consistency.
I love it when the conservatives all get together and take turns straw-manning the opposition and patting each other on the back.
“Like lightning striking” presumes both random and rare. If something happens all the time, it’s no longer “random” especially when it’s happening to people who are committing the same kind of action.
Right now there are far more white people getting away with crimes than black people (see my comment on another thread about drunk driving), so it’s much more likely that catching more criminals will probably result in the cohort of caught criminals having a racial makeup that is closer to that of the rest of the country.
I think the important thing to note is that basic moral decency requires us not to give half a fuck whether this is true or not.
Enforce the law, protect the innocent, and send the guilty to prison, and if we do it well enough we will probably find that there are fewer guilty people and we can probably imprison them for shorter terms.
Shouldn’t the racial make up of criminals caught resemble that of criminals period ?
It’ll be closer to that too.
Most homicides are intra-racial, but you seem to be holding the rate fix to get to your conclusions.
I don't think that's likely justified.
Just from opportunity (~5x as many white people as black), the inter-racial murder rates are probably skewed.
You assume there's one offender per victim. What if one group sees a higher rate of repeat offenders than the other? (I do not know the answer to this question and any guess on my part would be pure conjecture, but it's a possible explanation for the disconnect.)
I wonder if a grand bargain could be struck here to expand enforcement of financial and labor crimes at the same time.
Bargain with whom? Whose interest precisely is it that *any* kind of crime be under-enforced that it should be a “bargain” to enforce one or the other ?
The Republicans' No. 1 campaign promise is literally to stop tax enforcement, so ...
The claim is that it would be unfair to step up enforcement of the laws that marginalized people break, while white collar criminals and corporate wage thieves go unpunished. Sometimes extended to claim that because law-and-order sentiment focuses on marginalized-people crimes, it is in fact just animosity towards marginalized people and it a commitment to the rule of law in itself.
How on earth is it unfair to enforce the law? Lack of enforcement among marginalized groups is a particularly nasty social ill and injustice. I’d go as far as to say that the pro-crime (pseudo)-leftists are engaging in class warfare against the poor (who are— as we are obliged to recite nowadays—disproportionately poc)
To be fair, there has been lots of squawking from Republicans about the increase in enforcement personnel for the IRS, which is kinda analogous to the "defund the police" silliness from the left.
Who is the victim of tax evasion?
I don't think it's unfair to enforce the law; that's why I'm interested in political strategies that could make it happen. But the "pro crime" leftists are an important part of the electorate in a big city, and to get anything done you're going to have to address their fundamental concerns, like the (not unfounded) complaint that law-and-order, policing, and incarceration seem to be more about suppressing the poor than something also binding on the rich.
There is a great deal of evidence that “marginalized people” commit crimes far out of proportion to their population.
Is there any evidence whatsoever that there are appreciable numbers of white collar criminals or “corporate wage thieves” (whatever those are) are breaking the law with impunity?
My understanding is that "corporate wage thieves" are when people on hourly wages are forced (by threat of losing their job, not physical coercion) to do extra work "off the clock" for which they are not properly compensated. The comparison I once saw, which always strikes home to me, is that if a worker pulls $100 out of the cash register, puts it in his pocket, and walks away then that is considered a serious crime and he will be arrested. If an employer decides to simply not count $100 of a worker's labor and leave it out of his paycheck, the police usually consider that to be an employment dispute and not a criminal matter. Yet in both cases $100 was stolen.
I've seen that explanation before, but I tend to find it uncompelling because most, if not all, states have plaintiff friendly anti-wage theft laws to encourage civil enforcement (e.g., statutory minimum damages, doubling or trebling of the stolen amount, attorney fees, etc.) and which are certainly enough to make wage theft a poor business strategy.
Is there any evidence that the sort of thing you’re describing is not being suitably prosecuted?
Ok, I may have phrased it badly, but my point is rather the “grand bargain” framing as problematic. We all know that the republicans are traditionally there to help big money. But the phrasing here suggests the democrats are there to help violent criminals. That’s rather a rough rebuke.
The company I work for wants labor laws to be enforced because we want a level playing field.
Ken, what company to you work for, what industry is it in and what are you doing to help your state enforce their labor laws to level your playing field? Also, what are you doing to keep undocumented labor from undercutting documented labor? Probably nothing.
“…what are you doing to help your state enforce their labor laws to level your playing field?”
I am a member of an armed militia. We patrol the streets rain or shine.
Is that enough?
There’s an interesting tension in Matt’s thinking. He has talked about keeping things like prostitution and cocaine formally illegal but not enforcing such laws vigorously. The idea is to stigmatize and reduce such conduct without oodles of caging. Yet he also argues that swift and likely punishment is the best deterrent.
The basic problem is that a system designed to stamp out murder and robbery has been pressed into service to reduce vice crimes. This means that resources devoted to vice suppression inevitably come out of appropriations and manpower tasked to prevent violent crimes. At a typical felony arraignment calendar in Georgia, a third of the people are accused of victimless crimes. That’s a real problem when getting a fairly simple armed robbery case to trial takes over a year.
Punishing vice through civil penalties (possibly a portion of the offenders wealth/income) would be much better
Or just involuntarily commit addicts and people with debilitating mental illness to rehab and treatment. How is giving a large fine to a heroine addict with no money going to improve their problem?
I for one would not be comfortable with involuntarily committing someone to rehab without a significant due process procedure - aka a trial.
Portugal's system for substance abuse seems pretty good to me. We usually talk about things on a punitive / rehabilitative axis and looking through that lens it's probably a bit more punitive than many Americans imagine, but certainly rehablitative overall.
But to me the major difference in focus is accountability. Portugal's drug courts expect you to not abuse illegal drugs. They'll assist you in stopping if they can assist you, but if you continue to resist they'll increasingly use the stick alongside the carrot.
American policymakers seem to come at it from either a "more stick" or "more carrot" angle. I think we'd be better off thinking of it as "more accountability / more expectations". I wouldn't accept a family member of mine using heroin, for example, because I love them. I expect them to do better. We should love our fellow citizens enough to expect them not to abuse drugs.
Committing someone to rehab without them actually wanting to rehabilitate is pointless. All your really doing is forcing them into a different type of prison. That's a big step to me and deserves a noticeable hurdle for society to impose upon an individual.
The first two clauses of this seem like they're driven by the moral view expressed in the third rather than any sort of empiricism, given the data we have globally.
Also, as Wigan pointed out, there's a court system for this in Portugal and countries which follow that model. There *is* due process.
Making mental health treatment more widely available (like actual inpatient treatment for more than 10 days, and follow-up instead of just tossing patients out with a bunch of prescriptions) that might also prevent a lot of crime. A large number of drug addicts may be “self medicating” conditions that could be controlled with professional treatment.
Most users aren’t addicts with no money.
I really did mean substance addicts. Those people should be priority for enforcement of "vice" crimes. The lawyer who does a line of cocaine every few months isn't high priority.
Sure - but also, don't most significant substance addicts start off as occasional users who ramp up over time into what you see on the streets. So if we made people think that doing the initial dose was likely going to be caught and actually penalized then wouldn't it decrease the number of substance addicts over time?
But how likely are they to actually pay civil penalties? (I don't honestly know, my priors are "rather low," but not very strongly held when it comes to specifically issued citations, summons or tickets. Don't have any stats ready to hand).
you just involuntarily committed 10% of te population. good luck with that.
30 million Americans are debilitated (meaning seriously impaired in their ability to function in daily life) by severe mental illness and substance addiction? Seems high.
alcohol kills 300k americans a year, death seriously diminishes functionality, so yeah.
Compromise, and don’t involuntarily commit the dead.
Overall I agree with your approach. Less harsh sentences combined with a higher likelihood of getting caught if you commit a crime is the right way to go.
I think that increased survaillance is a reasonable punishment for committing a crime (and less cruel punishment than incarceration). However, I think it's worth discussing limits of surveillance. People who don't commit crimes rightly don't want to live in a surveillance state (like China) and are rightly concerned about issues of privacy and government control of daily life. So where do we draw the line? (I ask this as a genuine question, not a rhetorical one...as with most things, I'm sure there's a reasonable line to be drawn, and that those who use it as a rhetorical question usually are suffering from a slippery slope fallacy).
I was disturbed that the South Dakota program required abstinence from people on bond for dui who had not yet been convicted.
I’m also not sure that abstinence should be required in every first time drunk driving case. If there is an accident or the BAC is over 0.12 or there is a prior conviction, sure, but I don’t think someone driving home from a dinner with wine with a BAC of 0.09 should be subjected to the surveillance state.
I'm my personal experience .09 is actually like, pretty horribly intoxicated, but that gets to the real issue, which is that BAC is actually a trash proxy for impairment. Depending on the person, their tolerance, what they drank, when they drank it, what else they ate, their genetic predisposition, etc etc. Some people are sometimes trashed at like .04 and others are fine at .12. The breathalyzer system is really almost indefensible, even more so given how many of the impaired drivers out there are impaired by things other than alcohol.
Mandatory breathalyzer tests violate the 5th Amendment, but that ship sailed years ago.
How does it violate the 5th if its a condition of parole? You're free to say no to using it, but then you stay in prison and serve out your sentence...
I think Sean is saying in general--though the Supreme Court ruled long ago that it wasn't a violation.
But there's absolutely no problem having mandatory breathalyzer tests as a condition of parole.
I meant tests at random police stops. If someone is on parole they've already had due process, so at that point tests are fine.
But, driving isn't a right. You can make driving contingent on submitting to breathalyzer tests.
Mandatory breath tests do violate article i, section i, paragraph xvi of the Georgia constitution, but driving is a state issued license, not a right, so a license can be conditioned upon taking a breath test when there is probable cause.
Do they violate the constitution if the individual is on parole?
no, but they’d do a tribe test because those go back further
That seems like a better way to do it than using breathalyzer tests to put someone in prison
The constitutional amendment is basically dead. But if SCOTUS declared it a violation, I think you could get a breathalyzer exemption passed very quickly.
I'd imagine allowing breathalyzers would cross 95%+ in support. I spent time in a public defender's office and I'd support it.
Opposition would be limited to an extreme minority of people with very strong priors on restricting police power.
On international standards, a cutoff of 0.08 is on the high side. Where I live now has a cutoff of 0.05, and cutoffs of 0.03 or 0.05 are not uncommon elsewhere. (In Brazil, they allow no alcohol in your system...)
Anyway, I am fine with treating seriously driving at 0.09!
I've seen people argue that we should lower what BAC is unacceptable, but also make it a sliding scale of punishment. I.e. a .04-.07 is just a fine and you're not allowed to drive for the night. All the way up to, a .15 is an actual felony and not just a simple DUI. That way we can gradate the crime for how bad it actually is. If you're a .05 that's about as 'impaired' as driving sober but sleepy, which I think all of us have done at some point
I feel like this implies a far more linear relationship between BAC and impairment than is a reasonable reflection of reality.
Lol, I think maybe you're in the wrong place? Half this comment section are transit fanboyz/girlz, MY is widely known for his pro-transit takes, an the lone avidly pro-car guy gets flack all the time.
Just last week I said, "I want better transit partly so we can take away licenses for all sorts of offenses and people don't get to bitch" or something along those lines.
Yea, the problem with .08 is that it's basically a way of saying, "we want to charge people for reckless endangerment, but that's hard, so we'd rather implement a largely arbitrary cutoff that's easy to enforce in court but is actually poorly related to the thing we're actually trying to get at."
The problem with something like the .02 concept is that it it's largely unenforceable except as a post facto enhancer of other infractions because there's no outward evidence of impairment. It's also hard to treat "DUI" as a serious infraction when you're talking an .02-.03 cutoff.
My actual take is that we should be ditching the BaC infraction entirely and just making "driving above .08" an element of reckless endangerment statutes like it should be from the beginning. Send these people to jail for a crime, don't write meaningless tickets.
I have a completely opposite attitude: driving should be a rare and exceptional privilege for the 10-20% of people who are best at it, and the rest of us should have to rely on professional drivers for transportation.
Drinking and driving should be treated the way pilots who drink alcohol are treated - not allowed to touch the stuff eight hours before takeoff.
But I know, I'm weird and the US would have to completely restructure society to make that work.
"the US would have to completely restructure society to make that work."
I think this undersells the challenge dramatically! Have you ever lived in a non urban environment? I'm not even sure how someone who lives in a rural or exurban environment would make this work without truly enormous subsidies.
Good luck with that one...
How do you define who's best at it?
Complete disagreement. Bright line rules have real value. The BAC level says to people if you drink multiple beverages and drive an automobile you are violating the law.
It is much easier for people to understand compliance with the law.
And .08 is extremely high and shows how much we write the law to placate drinkers and drivers.
One of the craziest articles I've read in recent years was an Atlantic piece arguing that it was bad that people were using Ring cameras to catch porch pirates. https://www.theatlantic.com/technology/archive/2019/11/stealing-amazon-packages-age-nextdoor/598156/ Notably, the article isn't claiming that there is some bad second order effect, they just thought it was unfair that people who were minding their own business, stealing packages were getting arrested for it. I thought it was nuts. There are lots of things wrong with the criminal justice system, but arresting people for actually stealing isn't one of them.
"One man's package thief is another man's freedom fighter."
Recently had a package stolen, was very annoying.
Ok, but if you really think about it, all property is theft. So maybe you're the real criminal here.
If someone uses my stolen property, I just ask that they read an acknowledgment first.
No.
From evan bear's Twitter output, I'm pretty sure he's kidding.
If this were Twitter, you'd get a half dozen "This, but unironically" replies/quote tweets for this.
Both the "property is theft" and "taxation is theft" people are indeed out there in force on the internet.
My wife still talks about how someone stole an iPod Touch out of her Amazon package 13 years ago, and has decided that that neighborhood was a bad neighborhood based upon it.
There's also the (extremely weak imo) argument that sometimes the people on the Nest cam videos will be misidentified. Which is true, but I can't imagine that happens _more_ often than without literal video.
And then there's this, which makes me want to puke:
> Despite the much higher cost of white-collar crime, it seems to cause less societal hand-wringing than what might be caught on a Ring camera, said W. David Ball, a professor at Santa Clara University School of Law. “Did people really feel that crime was ‘out of control’ after Theranos?” he said. “People lost hundreds of millions of dollars. You would have to break into every single car in San Francisco for the next ten years to amount to the amount stolen under Theranos.”
I suppose it's *theoretically* stealing when you invest in a fraudulent company, but if you really expect anyone to get as worked up about a person with enough money to invest in a startup losing said money vs a person whose kid's birthday present was stolen off the porch by a drug addict....your priorities are bit skewed to say the least. And in any other context except this twisted one where we are not supposed to punish racial minorities for petty crime would you suggest that the mom who just ordered a package from amazon is a less deserving victim than some investor-class white person.
Great article! I read somewhere once (can't remember where and a Google search didn't come up with anything) that a large percent of crime in America is committed by a tiny percent of the population. The subject of the story was a one-woman crime wave and excellent anecdotal evidence of that thesis.
You clearly didn't read the whole article.
Not even 1/3 of the way into the article the subject admits to stealing "once or twice, three times at the most".
Of course there's a ton of evidence, including neighbors' mail she still had in her possession, that this was a big lie.
Having been caught stealing does tend to make people more suspicious of you in regard to stealing, much as having been caught doing all sorts of other crimes tends to make one suspect.
It also tends to damage your credibility when claiming that you definitely didn't steal that unopened package, addressed to someone else, that you are carrying home, but rather found it laying on the side of the road and assumed it was abandoned property, free for the taking.
How dare you!
Sure- so long as it applies to LEOs too. My philosophy is that if you want to be licensed by the state to do violence on its behalf, you're also subject to a much higher bar for conduct. Felons have to give their DNA samples? No problem- also, every single law enforcement officer should as well. (It would have caught the Golden State Killer much earlier!) Parolees are surveilled more heavily? Sounds good- and also let's make sure every single cop in America wears a body camera (Mass state troopers still don't, last time I checked), that they have to be on *at all times*, and that the footage is immediately owned & stored by a completely different government body. No warrant required to start an investigation for bad conduct, by becoming a cop you agree that your conduct is filmed 24/7/365 on duty. Every cop's face should be photographed and added to a facial recognition DB, so that they can be instantly IDed when they commit crimes in public.
LEOs should be required to testify against each other as a condition of employment or having a pension. Do you know of a fellow officer committing a crime on the job? You should be *required* to testify against them, on pain of possible arrest and losing your pension. Perjury on official paperwork should be publicly prosecuted. All LEO internal docs (emails, etc.) should be immediately FOIA-subject.
I'm pro law & order! Let's apply law and order to an organized gang well known for carrying out brutal crimes and then concealing them, the government
There are one or two quibbles I might pick here (while agreeing with the basic thrust overwhelmingly) but I'm 100% on board with "I'm a law and order person, and that's why those charged with upholding law and order by the state need to maintain exemplary conduct!"
Would you apply that same philosophy to other professions with the power of life and death? Should nurses and doctors be required to wear body cams?
Think about all the money saved in the courts if we had video of surgeons screwing up, and reacted immediately rather than spend hundreds of thousands trying to obfuscate around it.
I'm sympathetic to the argument, but I don't know that it actually works that way. There's a lot of context around a surgery or medical treatment, and many judgements that often have to be made by experts. And oftentimes the judgements of experts will disagree with each other.
More transparency might not hurt in terms of exposing real malpractice and nefarious behavior. But it also exposes every move to Monday-morning quarterbacking from people who may not know what they are talking about.
If you accept his logic as sound, I don't see any limiting principal on why this shouldn't apply to many major occupations:
1. Teachers: these individuals play a major role in shaping your values and life outcomes
2. Firefighters
3. Doctors/nurses: checkups are the difference between life and death
Not just checkups - every estimate of medical malpractice I've seen comes up with hundreds of thousands of deaths per year. It's crazy to be so worried about police malpractice but blaise about medical malpractice.
I think it comes down to a central fact. People are much more comfortable with surveillance of those they consider "others." Rightists are fine with their tough on crime platform because it isn't their neighborhood that gets the oversight.
Left of center police critics are fine with this surveillance of cops because they would never be one nor do they think much of the people who are cops.
That's actually 100% what I was thinking, too.
In both of your examples, I think more surveillance might be a good solution. But there's not a fair consideration given to the downsides to the people who would have to be surveyed. It's just like "F em", want to be a cop, deal with it".
You see this in attitudes toward teachers/police unions.
In both cases you have a coalition wanting to take away unionizing rights from a class of government workers.
In both cases they make good arguments that the services they provide are so important to society that unionizing is harmful. Police really can kill you. Teachers really can shape your life outcomes and your values.
But they only want to take unionizing rights away from the profession that belongs to the enemy tribe.
I think that we should just collect DNA from newborns -- I get that it has authoritarian overtones, but if you believe in liberal democracy and the rule of law, why not have a complete database going forward. It prevents all of the concerns about the bias in the criminal justice system meaning that some folks are more likely to have their DNA collected. There was a controversy in SF where a woman was prosecuted for some kind of property crime based on DNA from a rape kit she had provided in a previous investigation. Adding DNA collected in that way has obvious disincentives to reporting crime, but if we all knew that our DNA was in a database then we would all be on notice that it could be used against us in a criminal prosecution.
I really like the idea as a whole. I'd think a typical subpoena would handle requiring testimony against fellow cops. As for their own wrongdoing, are people allowed to give up their Fifth Amendment right against self-incrimination as a condition of employment? Even if not, forfeiture of a pension should do the trick.
I was in particular thinking of the shooting of Justine Damond in Milwaukee, where the shooting officer's fellow police refused to testify or say anything, and had to be subpoenaed by the prosecutor. They were- shocker- advised by the union not to cooperate with the investigation! That should be completely illegal- if you want to be authorized to do violence on behalf of the state, you should be required to 100% cooperate with every internal investigation.
'He said that his office started issuing subpoenas only after officers, acting on the union’s advice, turned down repeated requests to come in and provide statements.
“We tried again and again, and what they said was, ‘No, we’re not coming in.’ The Police Federation or its lawyers said, ‘No we shouldn’t,’ ” Freeman said'
https://www.startribune.com/tension-rises-between-county-attorney-police-union-in-noor-grand-jury-investigation/475458363/
Sounds good. Seeing as, as you note, the chief of the MPD was not willing to assist into an investigation of an officer shooting a completely innocent person, this just bolsters the common sense principal that a government agency should not be tasked with investigating itself. After all, if you heard about misconduct at the IRS or the NEA, would you trust them to investigate themselves?
So we can simply add this 'authority to order those officers to cooperate on pain of termination' to a civilian review board. And make sure that it's exempt from any union arbitration clause nonsense- if the officer declines to cooperate, they're fired immediately and no arbitrator can reinstate them, and they cannot be rehired for any LEO or government job anywhere in America. Plus, they lose 100% of pension accumulated to date
I am pretty confident that there's not one jurisdiction in the entire United States where a civilian run review board has the authority to compel to an LEO's testimony on pain of firing. (Not to mention that I explicitly included 'and this is exempt from arbitration', because I think we all know the government union arbitration game. Like 'America's worst cop', Florida LEO German Bosque- arrested 3 times, fired 7 times, he just keeps getting rehired thanks to arbitration!)
But I could certainly be wrong. I'm open to learning the name of even one municipality in America where a civilian review board can unilaterally fire LEOs. If you want to name even one....
From law enforcement to privacy to free speech and more, our policy mental models haven't caught up with or internalized the sea change in the facts of the world -- that we now live in a digital panopticon, where, like it or not, it's trivially easy to reconstruct what anyone was doing or saying, and where, from a collection of mostly privately owned but also public databases. Our mental models in these areas are still resisting accepting this fact, rather than accepting it as a given and the starting point that we should work with and use. For example, in a world of total surveillance and perfect enforcement, the more salient question about criminal laws become what they prohibit and what would be a reasonable (nearly certain) punishment. That's a different mentality than the "innocent until proven guilty"/"catch me if you can" crime deterrence model we've long had, a lottery model of very strict laws with draconian penalties, but poorly enforced.
I've wondered how traffic laws should work if everyone had a device on their car that recorded any time they were speeding. It's less obvious than it seems at first blush. We currently have a system that basically allows everyone to speed a little bit, but discourages you with reasonably large financial penalty if you get caught speeding more than reasonable. I don't think people would be OK with 100% enforcement, but I'm not sure what would replace it.
Speed limits conformed to road designs (and vice versa) could have a lot more legitimacy. When you build a highway and mark it at 30mph, then yeah, not okay with 100% enforcement.
I asked Matt a mailbox question about this that he unfortunately didn't answer. I'm wondering if car insurance companies will get there first by requiring transponders for their own bottom line desires. I'd also think you'd need universal transponders in order to make effective some other policies that are popular here, like congestion pricing, or replacing the gas tax with a per-mile tax.
You should change your handle to Ms. City of Trees.
I'm embarrassed that it even took me four seconds to get that, well done as always.
At least Progressive already has a "Snap Shot" program where they use information from your car's computer about when and how you drive to set your rates.
I've also thought about how EZ Pass could serve this function. Particularly on highways where you pass through multiple gates, it seems like simple math - if the speed limit is 65 mph, and you go through two gates 65 miles apart in less than an hour, you must, at some point, have been speeding.
Traffic laws need a radical revamp. We wrote laws for speed limits and stop signs and the like that are simple to state, but don’t actually do what we want. So in practice, people exceed speed limits on most cases but stay under in others; they yield at four way stops without stopping; etc. The Idaho stop law is one of the first attempts to fix this, but only for cyclists. With real speed detection we would need to fix the speed limit laws too.
Disagree as to stop signs. IME they get a pretty good approximation of compliance much more than they don't. It's one of the high-trust / non-defection behaviors that America is actually pretty good about engaging in.
Depends on the stop sign. 4-way stops are notorious for low compliance and also confusion as to who yields to who. Ideally, you replace those with roundabouts, which naturally obviates these problems, but unfortunately in developed areas they are often too cost prohibitive to construct.
Sorry to necropost, but the British-style mini-roundabout (which is just a painted white circle) doesn't need any physical work to construct.
It functions essentially as a four-way yield with really clear rules on who you yield to (you yield to the road to the right, or the road to the left in the US where you drive on the right). Most drivers drive over the painted circle, you're not supposed to make a massive effort to avoid it.
Hm, I'll have to look into this one more. The challenge would be making those rules really clear in a place where they are unfamiliar--especially because typical 4-way stop rules are to yield to the car to your *right* in right-side driving places.
Replace 4-way stops with 4-way yields, so people don't have to come to a complete stop before moving off, and the approximation of compliance would become actual compliance.
This is what I *want* to say, but then I realize that the law doesn't have a concept of "4-way yield". A "yield" sign means that the one leg of the intersection always yields to other legs, and you can't have that at all four legs of a four-way intersection. We need three distinct signs for full two-way stop, for full two-way yield, and for four-way yield.
I'm not sure that you've sat and watched at a four-way stop. If you actually look at the wheels of a car at a four-way stop, you'll see that only a very small fraction of vehicles actually come to a complete stop for even a moment. The vast majority slow down, let the vehicle with right-of-way through, and then speed up.
This is exactly the sort of behavior you can do at red lights legally, if you time things right - slow down as you approach so that you don't arrive until it's your turn to go, and save some momentum so you don't waste gas. But even though we think of red lights as somehow stricter than stop signs, the official rules allow this behavior at traffic lights and don't allow it at stop signs.
Two way stops, where one street has permanent right-of-way, and the other always yields, do tend to have a lot more actual stopping, but I think it's still probably only about 70-80% at most, when traffic isn't high on the main road.
People obey stop signs as much as they obey speed limits - they do a good job of following the behavior that is not quite the legally posted behavior, but is what the rule should have been.
Going slow and steady to preserve momentum is even more important when you're on a bike. That's why the Idaho stop is so awesome and sensible.
When I was a Boy Scout 60 years ago, I earned my safety merit badge standing two hours at a four-way stop recording how many cars came to a full stop. About 1% as I recall. Drivers generally have common sense. Which is why they also drive on highways at the speed of the traffic flow, even if it's 10 mph over the stated limit. They also generally obey the rule that at a four-way stop with two cars arriving at the same time, the driver on the right has right-of-way. I was recently at my wife's place in LeMans (France) where roundabouts have completely replaced four-way stops. The French know how to build them cheaply from the space in any intersection, and they greatly improve traffic flow. Bill DeBlasio, New York's unlamented ex-mayor, lowered NYC speed limits from 30 to 25 mph to reduce pedestrian deaths. No one abides by it.
I was recently driving in the south of France, and loved the roundabouts at every rural intersection! They also use them more at urban intersections than the US does, but after seeing it as universal at rural intersections, that clearly seems like the right way to go, rather than requiring a traffic light or a two way stop or a four-way stop, as we have all over the place in Texas!
Right, you'd want to relax speed limits on controlled access freeways (where the safest speed is often going the same as everyone else, and that tends to be about 10 MPH over what's posted), and make it stricter in places where there's a lot of non-motorist traffic.
Even better is to reconstruct streets to make it naturally prohibitive to break the posted speed limit, but that's a more expensive venture that takes more time.
I do a mix of agreeing and disagreeing about speed limits. Many drivers drive to the apparent speed encouraged by the road layout. In general, that should be how the speed limit is set, but there are cases where the speed limit is right and the road layout is giving the wrong messages to the driver.
There should probably be two sorts of limit signs, one that indicates "this is a normal 25mph or a normal 45mph, which is what we think you'd expect from the road" and another that says "look out, unusually low limit here; there's a reason for this, slow down; the road layout is lying to you". If the latter is rare enough that drivers will treat it as a hazard, then it would help a lot.
The second one should be, for instance, near elementary schools where small children are liable to wander out onto the street or where lots of kids arrive and leave at the start/end of the day, or where there is a blind bend just before a pedestrian crossing, or there's a hidden junction or whatever. But a combination of a hazard warning with a reduced limit would be a really useful sign.
There's a road near where I grew up which looked like it was an obvious 40/45mph road - straight, good visibility, no houses either side. But there was a bus stop one side and a pub about 50 yards in from the road the other side, and there were always people being hit when coming out of the pub to cross the road to the bus stop to get home. There isn't a hazard sign for "drunks like to cross the road here", but they dropped the limit to 30 just for that little stretch and installed cameras: far fewer people died. Lots of people who weren't locals hated this and complained because they thought the limit and camera were just to raise revenue.
Punishment is a wonderful thing. Not excessive punishment (definitionally bad) nor wrongful conviction (which isn’t punishment at all), but punishment fitting the crime, exacted by the state, after due process. It’s important for a basic moral reason: if someone has done wrong they should suffer the consequences. It’s also critical for a social reason. One of the key advances in the very existence of the state is the social contract in which individuals, families (and tribes/clans etc) surrender their claims to personal vendetta, even against the most heinous wrongs, and in return the state takes the responsibility to seek out and exact retribution on their behalf. This is hugely beneficial to society for a myriad of reasons. But we must not forget the hard basis of it all. The state has taken up the office of exact “vengeance” on our behalf. Thus punishment is an end in itself. Yes, we want to prevent recidivism. And yes, we want to deter others. But these are *additional* goals. They do not replace the basic point of punishment. So let’s not present punishment and enforcement as a zero sum game. In a well functioning society we have low crime rates, criminals are very likely to get caught, and then they are likely to get convicted and punished appropriately for the crimes they’ve committed. Its a sad state of affairs that MY feels the need to justify the idea of better enforcement per se, and that he presents laxer punishment as a carrot. Excessive punishment should never be tolerated, regardless of enforcement levels, nor should lax punishment.
Spot on.
"One of the key advances in the very existence of the state is the social contract in which individuals, families (and tribes/clans etc) surrender their claims to personal vendetta..."
Amerigo Bonasera (from The Godfather) nods in agreement
Your analysis doesn’t work for crimes that lack an individual victim. If Joe six pack drives drunk but isn’t that drunk and gets home safely, there is no vengeance to exact because no one got hurt.
Arguably there is - he put the safety of the public at risk. However I conceded in my original comment that this isn’t the *sole* reason for punishment, and it perhaps doesn’t apply in every case (that’s why we have additional, independent justifications, such as deterrence). However it is a foundational reason for the existence of criminal justice generally and must feature in every discussion related to the topic (or at least loom large in the background). I feel like MY’s post, like so many these days, are forgetting the basics.
I guess what I’m pushing back against is what I would call the dark side of puritanism— taking pleasure in punishing rule violations. This is a real human tendency, it is often called righteousness.
The state should be more like the temperate parent reluctantly spanking its children than the sadist who takes erotic pleasure in spanking.
“Punishment is a positive good in itself” is one step from getting off on spanking evildoers.
I think the parent analogy is totally false. The spanking is for the benefit of the child. Punishing criminals is not for their benefit but for that of society and the victims. I think it’s very healthy and human to take a measure of satisfaction or relief and yes, in certain cases even joy, in seeing Justice served: that Justice be served , and be *seen* served is crucial for society’s well being. However this response is of the victim and the people who care about them, and the public at large. The *application* of criminal Justice should not be emotion-driven and that’s why we have laws, professional judges and lawyers etc.
The problem with Puritanism as you put it is in over criminalizing things that ought not be criminalized and in being religious fanatics. a totally separate issue.
part of me agrees with you. this part also thinks public executions would be really cool if the condemned really were degenerate murderers.
a few well chosen executions per month could slack the public’s need for vengeance admirably. there could even me a democratic, participatory process to decide which death row inmates are actually executed.
i’m not sure this would work out well, but i’d be happy to try. it’s certainly a more elegant way of pacifying the demos than mass incarceration. i’d rather have public executions of murderers than long sentences for every 19 year old burglar
"a few well chosen executions per month could slack the public’s need for vengeance"
I don't think it works this way. I think if you whet the public's appetite with some public violence, they'll happily demand more to feed that appetite.
I agree. And one of the most immediate things one could do to implement it is move traffic enforcement and public nuisance events to non "police" officers (and to technology, including enlisting citizens to record parking infractions) so as to let "police" officers focus on crime deterrence. apprehension of suspects, and provision of evidence for prosecution.
Sending someone a ticket in the mail is not enforcement unless they actually pay the ticket. At some point, you need to be able to boot cars, garnish wages, arrest people, or otherwise create actual consequences, which hits up against the same problem of "some people are actually against enforcement."
bingo
But most people DO respond to mail fines so that frees resources for booting, garnishing, etc. In DC we have the additional problem that VA and MD do not honor our tickets (although I'd think that reports to credit bureaus of unpaid fines would be somewhat effective).
https://dcist.com/story/21/12/06/dc-council-cheh-henderson-department-public-works-ticketing-booting-cars/
"But DPW only has a staff of four right now operating in two booting teams, according to acting DPW Director Christine Davis. They can only boot about 50 vehicles a day, meaning it would take more than 25 years to get through a backlog at the current rate."
So we're only doing the sending tickets part -- in theory we could be freeing up resources, but that doesn't seem to be how this is playing out. Certainly MD and VA are a huge part of the problem, but it's not just them, and the fact that DC outlawed suspending licenses for unpaid tickets is also a part of the problem.
So each of the four booters is booting 12 a day. This sounds like a management/labor problem (lazy booters) or a technical problem (invent a better boot). I assume more booting will solve the MD/Va jurisdictional squabbles.
In the context of the vast amount of lack of enforcement of other traffic- and non-traffic related issues, I would conclude that while the mechanism for lack of enforcement may be something like "labor problem", the cause of the issue is "the city doesn't want to do enforcement."
Not suspending licenses for unpaid tickets is probably a mistake, but suspending licenses does not automatically get fines paid, either. Sounds like more booting would be a good use of the resources freed by more use of technology. And/or just an increase in resources; it ought to more than pay for itself. How about contracting with private towing companies?
There are all kinds of policies we can probably agree on! My point is that the same forces which drove lack of police enforcement are also driving lack of other enforcement, which is that the government does not want to be doing enforcement. This isn't a "if we just did this in a way other than cops" issue. We had police doing enforcement, and now we have nothing doing enforcement, unless you are the sort of person who pays your traffic tickets even if nothing is making you pay your traffic tickets. So the idea that if we do less police enforcement that will free up resources, or there's some technocratic fix here, while true in some sense is not true in the sense of actually being what happens.
Yes and no. Sure, no policy we might want to change just popped up with no politics behind it. But the whole point of Matt's essay is to try to change the politics, that, for example, we got specific non-enforcement bias from a misunderstanding of how best to reduce incarceration. If we think every issue is in a perfectly locked socio-political equilibrium, we are all wasting our time writing about it. :)
Just to hijack (carjack? :)) the conversation, technology exists to document and fine excessive muffler noise. It should be used but those non-"police," traffic officers. And although fines are not appropriate, excessive serene noise should also be reported to police, fire, and ambulance company headquarters.
"serene noise" is a wonderful typo ;)
and oxymoron!
It may not be oxymoronic in the future: when electric cars become more ubiquitous, there's a real concern of them being too quiet for non-motorists to detect and safely avoid.
That really is an easy problem to fix. You just put a piece of carboard on the frame so it makes noise as the spoked wheels turn, like we used to do with our bicycles. :)
Maybe they'll start looking up from their phones.
Largely agree, but if it is a repeat offense, fines probably are appropriate.
What is the alternative?
Getting buy in from their supervisors to take complaints seriously.
I mean, if we can't fine people who repeatedly violate excessive noise regulations, what alternative is there to stopping them?
Throwing them in jail? Confiscation of their noise-making device?
Yes, do fine them. And I'm not opposed to "police" officers stopping cars with (maybe more than one) outstanding violations.
I'm rejecting the idea that using technology to detect and send summands for traffic violation is useless. that only armed police officers can issue meaningful summons. Really, what's the difference between an officer-issued summands and a machine issued one? The person still has to pay by mail or show up in court?
Toll roads don't use officers in cars to issue tickets for not paying tolls.
He’s talking about sirens there. Maybe confiscating the sirens from the ambulance of a private hospital that is abusing them is appropriate, but it’s tantamount to putting them out of the emergency business.
Not necessarily. Sirens are OK. Not all are excessive. But they do not need to be audible3 and 4 block away for traffic to accommodate the emergency vehicle.
**PUBLIC** surveillance cameras, i.e. covering public spaces, are a good thing. The more of them the better. Nobody has a reasonable expectation of privacy in public.
Another way to think about it: if it would be OK for a cop to stand and look in the same direction as the camera, there's no privacy objection.
I can stand on the sidewalk and watch who comes in and out of your house, but a) you can see that I'm doing that and b) it's labor-intensive. If we were to move toward tech that could observe that cheaply and ubiquitously for everyone, I think that would have some dramatic implications, and I do not think that fears would (or should) be calmed just by saying "public space."
As I've said repeatedly in these comments the fundamental problem here is that we are over-criminalized and under-enforced
That is to say that our criminal code includes vast amounts of nannystate, social engineering horseshit like vice taxes and pretextual offenses and victimless crimes and municipal revenue generators that are not rightly criminal that create toxic incentives and vast opportunity for oppressive, discriminatory, corrupt and ineffectual enforcement.
We are also obviously under-enforced in that we don't reliably catch murderers and car thieves or intervene when some addict is ransacking the CVS, largely because policing primarily does those other things.
The problem with Matt's vast surveillance state is that he's regularly shown that he thinks the technocratic, nannystate, social engineering stuff is good actually. As long as people are committed to using the criminal code to manipulate the law abiding, instead of to reliably prevent and punish crime this sort of proposal should be opposed, tooth and nail, by anyone who values a free, pluralistic society. The threat posed by Matt's preference here is vast and while I agree that it offers effective enforcement opportunities, so long as the guardrails on WHAT these powers might be called upon to enforce are so obviously compromised it cannot be allowed.
I want to start off by saying that I miss Graham. I’d be quite curious to learn what he agrees and disagrees with in this article,
Where’d he go?
He said he was taking an indefinite hiatus, both on here and on his Graham Factor blog.
A criminologist friend of mine told me years ago that it is the odds of getting caught that had a demonstrated deterrent effect, not the consequences of getting caught. He was right.
Like most, I’m all in favor of catching more criminals but as things currently stand, that may not do much. In many areas, those caught for property crimes and even some assault and battery charges are out the same day they are caught (or the next). Prosecutors and police have given lower priority to such crimes due to various reasons, some political (requiring bail exacerbates the disproportionate racial effect) but also because of the sheer volume of such cases. There simply aren’t enough jail cells, guards, prosecutors or judges to increase the certainty of prosecution and conviction. Using all the technology that might lead to more arrests won’t necessarily result in increasing the deterrent effect of prosecution, conviction and incarceration. Unfortunately, bringing down crime rates is going to be much more difficult than taking DNA samples and enhancing surveillance.
Assuming criminal behavior is rationally motivated, increased apprehension ought to reduce crime by reducing the net benefit of the activity. However a substantial portion of crime is non-rational as a consequence of substance abuse among other things so cost-benefit calculations are irrelevant.
The bigger issue for me is that a pervasive surveillance society tied to facial recognition is subject to horrifying abuse. C.f. Xi's China. It is also true the the UK has had a pervasive CCT network without sliding into fascism. But the potential for abuse is clear. The only contemporary Republican whom I might trust with such a system is Mit Romney. There are also many Democratic officials who couldn't be trusted with it.
Fully rational crime would be deterred by the product of harshness of penalty and probability of getting caught. Purely irrational crime can only be deterred by understanding the weird irrational way a being operates. Most crime is partially rational - people are deterred by probability of getting caught, but aren’t particularly deterred differently by a one year sentence vs a thirty year sentence, or a thousand dollar fine vs a million dollar fine. Increasing probability of getting caught while proportionally decreasing punishment does generally increase the deterrent effect, though not to fully rational levels.
The way I've seen this expressed elsewhere indicates that, interestingly, criminals as a class are somewhat different in kind from most people: most humans are risk-averse but there's evidence that criminals are disproportionately risk-preferring.
What is the specific concern? What is it that the politicians would do?
Read 1984. Xi's China is already there.
I just think you're confusing the tools with the will-to-use-the-tools. If the US wants to imprison people for political activism or being an ethnic minority it can already do that. Likewise, China was locking up dissidents and Uyghurs long before they got ahold of facial recognition.
If the only thing Governments and police forces did was oppress people then I would support limiting their technological prowess. But the police also arrest people for crimes, and there are tens of thousands of fugitives out there, wanted for everything from rape to murder to kidnapping to just jumping bail. It would help our society to decrease that number.
And with an authoritarian movement gaining strength in this country, the will-to use-the-tools can be assumed. We are on the edge of a precipice. Expanding government surveillance powers at this time is inadvisable. I thought this was clearly implied in my original comment.
That's not what most people mean when they call Trump authoritarian. MAGA people want to do a lot of things, but setting up a dissident-squashing repressive surveillance state does not seem to be one of those things.