20 ms·

> Not only does this mean that police can very easily control crime statistics One could compare police statistics with victim surveys (https://en.wikipedia.or

> Not only does this mean that police can very easily control crime statistics One could compare police statistics with victim surveys (https://en.wikipedia.org/wiki/National_Crime_Victimization_Survey 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.

simion314 5y ago
For many crimes there might not be bodies to count. 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)
eesmith 5y ago
Indeed, it opens up the topic to a wider question - why are some things counted as crimes and others not? 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 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_Litigation 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".
- chuckee 5y ago
  A moment ago the conversation was about violent crime and "over-policing creates racist bias in crime data". Now it's about laws treating corporate crimes more leniently. 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.
  - geofft 5y ago
    I think the wage theft point is relevant (I was going to bring it up too) for two reasons: 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-and-nonfatal-violent-victimization-2008-2012 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.
  - eesmith 5y ago
    And you've deftly moved past the question "How do you determine if the police are being successful?" 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.

simion314 5y ago

For many crimes there might not be bodies to count. 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)

eesmith 5y ago

Indeed, it opens up the topic to a wider question - why are some things counted as crimes and others not? 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 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_Litigation 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".

chuckee 5y ago
A moment ago the conversation was about violent crime and "over-policing creates racist bias in crime data". Now it's about laws treating corporate crimes more leniently. 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.
- geofft 5y ago
  I think the wage theft point is relevant (I was going to bring it up too) for two reasons: 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-and-nonfatal-violent-victimization-2008-2012 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.
- eesmith 5y ago
  And you've deftly moved past the question "How do you determine if the police are being successful?" 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.