Predictive policing meant that more black people were arrested than white people. A new system, DICFP, was introduced.
DICFP means that more black people were arrested than white people.
Maybe the problem isn't the way the police are operating.
What is the problem then? :)
Or that specifically making things illegal because they are popular in a subculture can over decades drive those subcultures to integrate an inclination to crime in general into the subculture?
We can't effect change if we aren't willing to talk about the causes for the problem, or even acknowledge the existence of a problem.
They’re a uniquely violent demographic and it’s unsurprising that their arrest statistics don’t match other demographics.
The statistical differences are a result of that underlying difference in rate — rather than anything about the police. No change in police policy except ignoring violent felonies committed by blacks can fix the underlying per capita imbalance.
And no — that’s not a race thing, as shown by recent African migrants and other immigrants. It’s a cyclic poverty and crime thing… but that doesn’t change that there’s a correlation between violent felonies and race.
A correlation that gets carried forward into arrest statistics.
To my knowledge there’s no dataset on who actually does crimes regardless of police involvement (eg: including noninvestigated crimes)
What's more sad is that this is clearly fueling racism of the worst kind - I call it pseudo-intellectual racism. It's not the kind born of ignorance and cultural differences. It's racism that was reached through broken reasoning, resulting in one who believes their racism is justified, and thus, isn't racism.
Let me put it this way: would you willingly move to the ghetto, since as you imply these are friendly people just like you and me who are merely victims of disproportionate scrutiny by police?
Modern antiracism is a competition in mental gynmastics, built entirely upon denial of statistics and science. If you are going to acknowledge differences in culture then logical consistency requires acknowledging that certain cultures are more violent and criminal on average than others. That's the nature of our reality and to deny this fact is naive.
That's slimy, anti-intellectual reasoning. The fact that an opinion is unpopular with the mob does not imply that it is wrong.
In any case at least we've moved past the nonsensical assertion that crime rates only appear to be higher because of overpolicing.
200k+ years of divergent evolution, adaptation to drastically different environments, including geographically isolated interbreeding with other non-human homonids, lead to the same genetic differences that allow for accurate and precise geneologic inference from genetic samples, including the unique genetic makeup of african americans relative to other subsaharan african populations.
I'm not denying that history plays a role in the african american's plight. But you are completely dismissing the unpleasant implications of scientifically documented reality, blindly conflating the immoral nature of racial discrimination with the blatantly pseudo-scientific idea that genetic makeup does not influence outcomes on social scales, or worse, buying into the completely nonsensical idea that race doesn't exist. That's some 60 years of "one race human race" propaganda that you have heard for your entire life, so I don't entirely blame you for your naivety.
But evolution did not magically stop above the shoulders. The massive physiological differences - skeletal structure, musculature, heart/lung size, blood differences, body proportion, facial structure - are obvious. But although blindly virtuous individuals like yourself make it social suicide to discuss openly, these same differences are also quite convenient explanations for the variations in average achievement (and criminality) among various genetically distinct demographics. In fact we are aware of hundreds of specific genes, and their relative abundances, which are strongly correlated/anticorrelated with intelligence. Moreover we know of dozens of specific genes which directly and indirectly influence behavior, including the so called warrior gene which has been indisputably shown to completely change the metabolism of monomanines which directly influence behavior, including aggression.
Whether you believe that something is racist has literally zero bearing on whether or not it is a scientific fact. And architecting a society while blindly denying such demonstrated truths can only lead to suboptimal outcomes for all of stakeholders. Don't worry, none of this is an excuse to discriminate against individuals, because we are talking about statistical distributions.
But failing to acknowledge that such genetic differences have at least some degree of influence over disparate outcomes at scale in a perfect meritocracy is pure anti-intellectualism, and I am no coward for standing up to the woke mob.
Poverty + racism + the powers that be identifying black people as "super predators" and touting that they are more likely to arrest and more harshly punish black people for crimes != black people are criminal.
The system nurtures poorer black people to be more criminal than white people. It's not the nature of any group of people to be criminal or as a rule to intentionally flout societal norms.
There is no genetic component of that scenario.
Therefore, it is not that black people are fundamentally bad and unfixable. Any failings that you can find inexorably lead back to the flaws of the system itself.
If it is that the system itself is flawed and moldy and in desperate need of repair, and the people who refuse to so much as ponder whether there could be an explanation other than "black people are bad", well, it is safe to say that those people are in thrall to the powers that be, servants to the status quo; they are people who do no directly racist thing and yet, as if by a dark satanic miracle, are racist due to their own willful ignorance.
> genetic predisposition to lower intelligence and/or aggression
> certain cultures are more violent and criminal
Anti-racism is a competition against your mentality. Why put in so much effort to make things harder for others?
The problems are multi-faceted:
- some crimes, especially relating to drugs, were specifically invented (or their sentencing made harsher) to target people of color. Just remember the infamous quote of Nixon's top advisor: "We could arrest their leaders. raid their homes, break up their meetings, and vilify them night after night on the evening news. Did we know we were lying about the drugs? Of course we did."
- poverty, drug abuse and crime form a vicious circle. Over decades, that was mostly restricted to people of color, but the opioid crisis showed that the very same relationship also holds up in White, former solidly middle class people.
- related: prison sentences, even short stints in jail, for bullshit charges can and will have disastrous consequences like loss of employment and housing, which combined with the complete lack of a social security network leaves many people with only crime as alternative to make ends meet.
- "predictive policing", similar to AI/ML technology, feeds on existing biases in the datasets. When the "classic way" shows a high rate of crime in a PoC and gang dominated neighborhood, any form of "prediction" will direct more police presence there than to White, "rich" neighborhoods.
- related: where there is more police, the police will also find more crime. To take the drugs example: two people exchanging cash for crack on the street or an old, run-down car are visible, easy targets for cops. A "cocaine taxi" delivering stuff in an upper-class 'hood, with a brand new 7 series BMW? Won't ever get stopped for a random traffic control, despite the crime being the exact same.
Crime can't be combatted solely by using police repression. Anyone who seriously wants to tackle it has to go deep to the roots: a lack of societally positive perspective for way too many people of color thanks to structural discrimination, toxin exposure in "bad 'hoods" (lead paint, industrial effluents, traffic-related stuff like particulate matter dust), quality of education in public schools, a lack of clean and safe drugs forcing people to street vendors, un(der)educated police officers, ... the list is endless, and to fight the problems requires immense amounts of money and political clout.
You're proposing to pull the police from the bad or black areas and have them police rich or white neighborhoods?
This smells like the white savior complex, no offense. A safe person from behind the keyboard, wants police removed from bad black areas...even though the people in those areas want more police
https://news.gallup.com/poll/316571/black-americans-police-r...
A lot of people break down the problems into a series of axioms or beliefs instead of systems of inputs and outputs. So while policing isn’t enough to solve the underlying problem, removing policing makes the underlying problems worse.
Worth noting this is disputed.
Doesn't make the point itself invalid though - just look at punishment disparities between crack cocaine and ordinary cocaine (https://www.aclu.org/other/cracks-system-20-years-unjust-fed...).
How do you determine if the police are being successful? If there are more arrests, are they more successful? Perhaps you ask about crime statistics - but those statistics are recorded by the police. I think we all have anecdotal stories of police refusing to take police reports or saying that something is a civil matter, or simply witnessing violations of minor laws (jaywalking, littering, etc.) and doing nothing. Not only does this mean that police can very easily control crime statistics, it means that if your actual goal is preventing some specific crimes, it's not guaranteed that the police are actually working on it.
One theory is that the function of the police is to maintain order, by which is meant the status quo. Police exist to make sure that white neighborhoods don't start to feel like black neighborhoods, and that black neighborhoods don't start to feel like white neighborhoods. In that model, the two data-driven approaches (predictive policing and DICFP) fall into the same trap that any hacker on HN who has run into office politics is familiar with: the data isn't being used to decide what to do, it's being used to defend an existing decision by powerful people and argue that the decision is right. Should we keep text-based content or pivot to video? Facebook's PMs, who keep the official record of whether people watch videos, have data claiming we should pivot to video. Should we keep patrolling the same neighborhoods we've always patrolled? The police, who keep the official record of crimes, have data saying that crime is bad there.
One could compare police statistics with victim surveys (https://en.wikipedia.org/wiki/National_Crime_Victimization_S...) to check how close they match. Or look at homicide statistics, that are not so easy to control.
But then one could no longer dismiss all inconvenient data with a simple "the police are racist". So one keeps one's head in the sand and sticks to that line.
For example, why isn't wage theft included in the NCVS?
Surface-level answer: because in most jurisdictions, it's not a criminal offense. Steal $100 from your employer? Go to jail. Your employer requires unpaid overtime adding up to $1,000? A slap on the wrist at best.
Yet employees are victims to billions of dollars of wage theft every year! "Sixty-eight percent of the surveyed workers experienced at least one pay-related violation in the week prior to the survey. On average the workers in the three cities lost a total of $2,634 annually due to workplace violations, out of an average income of $17,616, which translates into wage theft of fifteen percent of income." - https://en.wikipedia.org/wiki/Wage_theft . Does the NCVS list that over half of employees are victims of wage theft?
Why, it's almost as if the laws were structured against poor people in favor of business. Hmm, and ethnic minorities are overrepresented in the poor population.
What about the illegal collusion between Silicon Valley companies to suppress worker wages? https://en.wikipedia.org/wiki/High-Tech_Employee_Antitrust_L... . Were all 64,000 employees counted as victims in the NCVS? Again, no, because that was a civil suit, not a crime that the police should be involved in.
But again, why is it under civil law, not criminal? (There's no hint that this wage theft is due to racism.)
Or, were the events leading to the lead poisoning of the water of Flint, MI and its coverups a crime? If so, why isn't the population of Flint included as victims? If not, why not?
Or, in discussions related to the 'school-prison nexus', when disproportionate disciplinary actions are taken against students of color, are those students counted as victims? Why or why not?
You are correct. All this inconvenient data indeed makes people dismiss the 'simple "the police are racist"' and replace it with "the system is both racist and biased in favor of the rich".
You raise good points, and use them to deftly move past all the claims of racism that homicide statistics and victim surveys largely (but not entirely) refute.
Is your metric only homicide statistics?
geofft commented "police refusing to take police reports or saying that something is a civil matter". My examples were meant to reinforce that issue, where crime statistics place far less weight on theft of services than theft of property, even when the monetary amounts involved are similar.
FWIW, wage theft isn't only a corporate crime. If I promise to pay someone to mow my lawn, and he does it, and I don't pay him (perhaps threatening to call the police because I suspect he's an immigrant without work papers), then that's also wage theft, even though I am not a corporation.
You're welcome to continue onto ever more abstract or systemic claims of racism and bias, but please don't begrudge me pointing out when some of the stepping stones crumble along the way.
On could also compare other statistics, and show that police are over-funded compared to other sorts of enforcement agencies.
The very act of picking which statistics to use is itself biased.
Simply pointing to NCVS is not going to be convincing. Which numbers should we look at? What correlations should we draw? How should we understand the biases in the interpretation? What alternative explanations are there for whatever it is you haven't actually said?
Without pointers to what what your actual point is, I figured it was just as appropriate to point out how your foundation is itself crumbled.
1. Directly, wage theft affects people's ability to get out of poverty and build wealth. Wage theft is a problem that disproportionately affects lower-income people (because they're more likely to rely on wages, instead of salaries or investments) and is disproportionately worse for lower-wealth people (because they can't as easily absorb a loss of $1,000 by covering it out of savings). We know, from those very same victim surveys, that poverty correlates with crime. In fact https://bjs.ojp.gov/library/publications/household-poverty-a... points out that poverty correlates with being a victim of crime.
2. Less directly, but more interestingly from a philosophical standpoint, wage theft is a great example of a crime where how you classify it impacts statistics. Victim surveys tend to look at robberies. Wage theft isn't a "robbery," but if you lose $1,000 from your employer, that's much worse for you than being mugged at knife-point and losing $100 from your wallet. And if you look at homicides: Los Angeles has 200-something homicides per year. It also has 40-something police-caused deaths per year, which are generally not called "homicides." It also has 200-something deaths from car accidents a year, which are also generally not called "homicides." You're equally dead through any of these means, but crime surveys create a specific definition of "violent crime" that only looks at certain forms of violence.
Because the two are not at all the same.
I would imagine the reason is that everything should already be on record, and you'd be expected to take said employer to court. That said, if employer tried to explicitly commit fraud e.g. forge documents, illegal employment, I'd expect someone to see jail time.
Theft, on the other hand, is inherently off-the-record and fraudulent.
Also, if said employer forced you to work I'd expect to see punishment vs asking (possibly illegally) for you to work overtime - in the case of theft no-one gets the right to refuse.
> But again, why is it under civil law, not criminal?
NAL but I think its something to do with breaking/violating contracts versus the law. It makes sense to me. You seem to be focusing on the sums involved rather than the context of the theft.
> it's almost as if the laws were structured against poor people in favor of business
legal action is expected (in the US) to be funded by the participants, i.e. they put in a stake. This means those with less money will always have fewer options. A small business can also suffer this way.
The flip side is those with wealth have more to lose, and big pockets often attract lawsuits.
What we perhaps need is better union - no-win-no-fee lawyers to help combat wage theft, but I suspect the cost of layers now is so high that lower salaries just aren't worth the fight - 2634 USD pa might not cover a lot, so you'd need to pool resources and make it all v efficient. I'd be interested in what barriers there are to this.
> lead poisoning of the water of Flint, MI and its coverups a crime?
I think it should be, absolutely. The whole think with Depont dumping shit in rivers is also terrible. I wish the government were more pro-active in punishing these actions, and I'm quite frankly surprised how little action there is - a sign of corruption at a higher level?
Which, as you seem to concur, makes it hard for low-wage workers to fight against wage theft.
> Theft, on the other hand, is inherently off-the-record and fraudulent.
The Wikipedia entry for wage theft links to https://www.npr.org/2013/04/10/176677299/construction-boomin... with an example of an employer not paying $1,200 in wages until the employee went to the Texas Workforce Commission.
Why should theft of $1,200 cash be a criminal act, while theft of $1,200 worth of services is not?
I don't see how this isn't fraudulent. There's no right to refuse after you've already done the work. And the employee must have originally arranged the work off-the-record because you don't tell the government you had an employee you paid $0.
The Wikipedia article points to other types of wage theft, like employee misclassification, which seems very much like either forging documents or illegal employment.
> I'd expect someone to see jail time
And if they don't, what do you expect to happen?
Because wage theft is rarely punished at all, much less resulting in jail time.
> legal action is expected (in the US) to be funded by the participants
In civil law, yes. Not criminal law.
Why is wage theft not part of criminal law (excepting a few jurisdictions), even for those cases where the employer is doing things off-the-record, fraudulently, and the employee has no right to refuse?
Even within civil law, why is the civil liability to employer so low? There's less need to pool resources if wage theft had treble damages.
You implicitly agree to take on a debt when you don't ask for payment upfront, but either way you still should have a contract of some sort. Just because you call it theft doesn't make it so - failure to deliver on a promise (contract) doesn't retroactively become theft.
> the employee must have originally arranged the work off-the-record
Illegally?.. You don't get the same protection of law if you conspire to stiff the taxman.
> And if they don't, what do you expect to happen? Because wage theft..
My comment wrt expectation was about fraud, not "wage theft".
> even for those cases where the employer is doing things off-the-record, fraudulently, and the employee has no right to refuse?
off-the-record and fraudulent activity should be punished, but I see no evidence it is being ignored. If employees are also complicit, maybe that's the reason these cases are not heard.
> why is the civil liability to employer so low?
presumably b/c, without fraud, there should be all the evidence/materials to prove wrong doing in the normal manner (i.e. in court). I agree, though, that those who do it blatantly should be additionally punished if it seems they are banking on poor organisation/resources of their employees.
Why are the enforcement agencies responsible for preventing wage theft - including "the taxman" - so underfunded as compared to the police, even when we're talking about comparable economic effects in the billions of dollars?
When talking about correlations between race and being a "victim", why should we only (or primarily) consider criminal statistics?
Is there a national equivalent of the National Crime Victimization Survey which covers wage theft, workplace harassment, and other illegal activities which aren't part of criminal law?
I dunno. Me code.
The taxman isn't formally concerned with wage theft, they are concerned with (tax) fraud. I'm not sure what other enforcement agencies there would be - the Department of Labour?
> When talking about correlations between race and being a "victim", why should we only (or primarily) consider criminal statistics?
We shouldn't, but that's a generalisation of the original question: why is wage theft not treated like a criminal offence.
Also, correlations with race doesn't imply causation: If race correlates with X then anything that correlates with X will also correlate with race. Race correlates with wealth, so anything that correlates with wealth can be spuriously correlated to race. If you normalise for wealth and the racial correlation disappears (I do not know if this is the case) then you have a stronger case for a (primary) cause.
> which covers wage theft, workplace harassment
presumably court cases and lawsuit in non-criminal cases. and other non-legal things such as polls and complaints. But I'm not sure why you expand this to workplace harassment etc - legal resources should be better provided at the lower end, but the courts are the proper place for these civil issues, not criminalising (recategorising) everything for the sake of fixing a funding problem.
So I think that just from a math point of view(no race/social stuff interpretation) , you can be in a situation where the starting data combined with bad algorithm will keep you stuck in a local minimum. Probably math people can model this issue with abstract terms not to offend any color of politics you prefer but you have the issue of bad implementation and as always garbage in is garbage out, so you might need an algorithm that starts with zero input and uses random data and converges slowly , but you still need a way to make sure you keep the algorithm evolving with changes over large time (maybe reset it and make it start over after a few years)
If I live in Walnut Creek I don’t care what happens on international boulevard in Oakland. If they pull all police it does not affect me much. It will affect the locals. Just as pulling police in WC would be a negative for me if I lived in Walnut Creek.
Even poor people do not want to live in crime zones but some people believe they be better are off living with crime because statistical policing is worse than no policing. I would disagree.
You should get an athletics medal for that leap.
One of the things:
How safe do you feel going from e.g pub/cinema/blabla to your home at 01:00?
It's not perfect proxy, of course.
This is an extremely low bar to clear. At least in these parts of the world street mugging is very uncommon. Yet, it has little to do with police work or presence.