This site makes extensive use of JavaScript.
Please enable JavaScript in your browser.
Live
PTR
10.2.7
PTR
10.2.6
Beta
Racism
Post Reply
Return to board index
Post by
Skithus
Here is an interesting thing to look at, its from the NYPD's 2008 end of year enforcement report.
Misdemeanor
Criminal Mischief
The first number of what percentage of the VICTIMS where of that race
The Second number is what percentage of the SUSPECTS were of that race
And the Third number is what percentage of ARRESTEES were of that race
BLACK 33.9% 50.3% 35.1%
WHITE 34.3% 17.6% 21.3%
HISPANIC 23.3% 29.0% 40.9%
So between black and white, we've got a very close distribution of victims, implying that in the case of a Misdemeanor the victim is basically equally likely to be either race.
We can also see that Black's are ahead of Whites on actual arrests, but behind Hispanics.
The interesting thing is that Black's are FAR FAR ahead on the suspects list.
What this tells us, is that in the case of a Misdemeanor crime NYC despite the fact that there is a higher likelihood the crime was committed by a Hispanic, Black's are the prime suspects.
I think that's a fairly clear case illustrating Prejudice (I'll informed Prejudice at that) In the police force.
Source
Post by
MyTie
Here is an interesting thing to look at, its from the NYPD's 2008 end of year enforcement report.
Misdemeanor
Criminal Mischief
The first number of what percentage of the VICTIMS where of that race
The Second number is what percentage of the SUSPECTS were of that race
And the Third number is what percentage of ARRESTEES were of that race
BLACK 33.9% 50.3% 35.1%
WHITE 34.3% 17.6% 21.3%
HISPANIC 23.3% 29.0% 40.9%
So between black and white, we've got a very close distribution of victims, imply that in the case of a Misdemeanor the victim is basically equally likely to be either race.
We can also see that Black's are slightly ahead of Whites on actual arrests, but behind Hispanics.
The interesting thing is that Black's are FAR FAR ahead on the suspects list.
What this tells us, is that in the case of a Misdemeanor crime NYC despite the fact that there is a higher likelihood the crime was committed by a Hispanic, Black's are the prime suspects.
I think that's a fairly clear case illustrating Prejudice (I'll informed Prejudice at that) In the police force.
Source
Great point. I'd be interested in knowing if the suspect status is earned because of race, or because the suspect actually perpetrated a crime.
Post by
Skithus
Here is an interesting thing to look at, its from the NYPD's 2008 end of year enforcement report.
Misdemeanor
Criminal Mischief
The first number of what percentage of the VICTIMS where of that race
The Second number is what percentage of the SUSPECTS were of that race
And the Third number is what percentage of ARRESTEES were of that race
BLACK 33.9% 50.3% 35.1%
WHITE 34.3% 17.6% 21.3%
HISPANIC 23.3% 29.0% 40.9%
So between black and white, we've got a very close distribution of victims, implying that in the case of a Misdemeanor the victim is basically equally likely to be either race.
We can also see that Black's are ahead of Whites on actual arrests, but behind Hispanics.
The interesting thing is that Black's are FAR FAR ahead on the suspects list.
What this tells us, is that in the case of a Misdemeanor crime NYC despite the fact that there is a higher likelihood the crime was committed by a Hispanic, Black's are the prime suspects.
I think that's a fairly clear case illustrating Prejudice (I'll informed Prejudice at that) In the police force.
Source
Great point. I'd be interested in knowing if the suspect status is earned because of race, or because the suspect actually perpetrated a crime.
If by perpetrated a crime you mean they have a past offense, the numbers still wouldn't add up unless the police believed that a select few Black's were frequent re-offenders. And even then, given the lower actual arrest percentage They'd still be incorrect.
There is still a radically unequal distribution of suspicion on blacks. Whites and Hispanics are mathematically more equal, as In Whites commit ~20% of crimes, and are suspects ~15% of the time. Hispanics Commit twice as many crimes at ~40% but are suspects around twice as often at ~30%
That makes mathematical sense, if a group commits twice the crimes, then suspecting them twice as often is logical.
The only scenario I can see this math bear-out other then the Police are Prejudiced against Blacks more then Hispanics, is that Blacks are better at not actually getting arrested. As in the police were ultimately unable to pin the crime on the actual criminal.
Post by
MyTie
If by perpetrated a crime you mean they have a past offense, the numbers still wouldn't add up unless the police believed that a select few Black's were frequent re-offenders. And even then, given the lower actual arrest percentage They'd still be incorrect.
The only scenario I can see this math bear-out other then the Police are Prejudiced against Blacks more then Hispanics, is that Blacks are better at not actually getting arrested. As in the police were ultimately unable to pin the crime on the actual criminal.
That's all quite possible, but not entirely self evident. For instance, it doesn't take into account that one arrest may account for a suspect in several crimes (ie multiple offender). So, someone may be a suspect 10 times, and have 1 arrest.
I actually worked in law enforcement, as a dispatcher for the Sheriff's Department, and I know there are strict guidelines for investigation, and that most Police Officers don't just go on hunches, but most criminals drop themselves into the Officer's laps, usually through stupidity. Smart criminals get away or require years of investigation. Due to this, I don't think that racism is accountable for such a vast pool of suspects. I'm not saying that it doesn't play a role, especially with your occasional poor police officer. I'd just be interested in knowing how what these statistics really reflect.
Post by
Skithus
An initial crime complaint report may record a crime with several suspects some of whom may be arrested at the time the crime is reported and recorded. The non-arrested suspects are included in the “suspect” statistics in the tables and charts included in this report and arrested individuals are included in the “arrestee” statistics. At the time these statistics were compiled additional follow up arrests may have been made increasing the number of arrested suspects. In so far as that situation has occurred within the 2008 data presented the suspect and arrests data will overlap and count the same individual as both suspect and arrestee if a follow up arrest has been made during the period.
That's from the source, explaining how they arrive at the suspects number.
Post by
168916
This post was from a user who has deleted their account.
Post by
MyTie
Ok. So you are saying it isn't effective. If you agree that "put statistical formula" here is true, you still don't agree that racial profiling is effective?
If A is more statistically likely than B to be X, and you choose A, you should be more likely to get X. However, you don't feel that employing this formula will lead to a finding of X?
Post by
168916
This post was from a user who has deleted their account.
Post by
MyTie
I agree that other factors must be taken into consideration, but we are talking about other factors being equal, as we are focusing this discussion on race, not what clothing people wear. As for race not being "causal" toward criminality, I would agree, but that doesn't mean that race isn't an indicator of criminality. A black person who commits a crime doesn't commit crimes because he is black, but because of other factors. I feel that these factors may come from his upbringing, or societal factors, etc. However, the cause of why blacks are more statistically likely to commit crimes is another subject entirely.
Blacks are more likely to commit crimes than whites in NYC.
All other factors equal, a random black person is more likely to be a criminal than a random white person.
Based on that, why does considering race
among other factors
when profiling a person not produce more
effective
profiling? They do it all the time when profiling serial killers, because serial killers are usually white. They don't rule out other races, but they do consider the overwhelming odds that the criminal is white.
Post by
Orranis
I also found it interesting that whites were the only people that were more likely to be arrested than suspected.
Post by
168916
This post was from a user who has deleted their account.
Post by
MyTie
Race is an indicator of criminality if there's a correlation, and it's an effective indicator to profile on if doing so increases predictive abilityWhat do you mean "increases predictive ability"?
Post by
ElhonnaDS
I think a point that people seem to not be considering, on the effectiveness argument, that when dealing with stopping random people of x ethnicity to find criminals, the percentage of criminals that are of that ethnicity is far less relevant than the percentage of that ethnicity than are criminals.
For example, males are much, much more likely to be rapists than women, based on statistics of convicted rapists. The incident of female rapists is so low, in fact, that I couldn't even find numbers, so if we say that 95% are male we're probably really over estimating how many are female.
Based on your logic, that it makes sense to stop people based on them statistically being more represented in a criminal group, the argument is much stronger to stop men randomly to check if they're carrying rape equipment, or show signs of a struggle or recent intercourse, than it is to stop specific ethnicities for evidence of criminal activity, because the percentage of people who commit rape being men is much higher.
The reason this would be grossly ineffective, is that you have to take into consideration what percentage of men are rapists. If one man in every 10,000 are rapists, then you'd be spending the manpower to stop and search 10,000 men before this caught someone. And that manpower, if it were focused on more specific indicators, or causal indicators, could probably be much more effective at stopping crime, than by randomly searching every man since rapists are more likely to be men.
The same thing applies to racial profiling. Don't show me statistics about what percentage of criminals are of each ethnicity. Show me statistics on what percentage of each ethnicity are criminals. If it's not a significant percentage- which I imagine it wouldn't be- then think about how much time and manpower would have to be output statistically before you caught criminals through racial profiling, and how much better those resources could be applied when looking for directly related factors instead on coincidental factors.
I would say that you'd be right in saying that a high percentage of people who are in a gang are criminals, and so we should check people out who wear known gang colors. I'd say a high percentage of people who are lurking about on commercial property late at night, after it's closed, are likely to be criminals. I'd say that if a specific building has known drug activity, and people visit that building regularly, there'd be a high enough percentage of them that were involved, to make searching people based on that effective.
However, for racial profiling to be effective, you'd have to show me that a large enough percentage of people of that ethnicity are criminals that it would turn out significant results to do that, and I don't think those numbers are there. If they are, please correct me.
And all of that is aside from the moral implications of racial profiling, or what it costs in terms of perpetuating attitudes and resentment in the country.
Post by
MyTie
I think a point that people seem to not be considering, on the effectiveness argument, that when dealing with stopping random people of x ethnicity to find criminals, the percentage of criminals that are of that ethnicity is far less relevant than the percentage of that ethnicity than are criminals.
For example, males are much, much more likely to be rapists than women, based on statistics of convicted rapists. The incident of female rapists is so low, in fact, that I couldn't even find numbers, so if we say that 95% are male we're probably really over estimating how many are female.
Based on your logic, that it makes sense to stop people based on them statistically being more represented in a criminal group, the argument is much stronger to stop men randomly to check if they're carrying rape equipment, or show signs of a struggle or recent intercourse, than it is to stop specific ethnicities for evidence of criminal activity, because the percentage of people who commit rape being men is much higher.
The reason this would be grossly ineffective, is that you have to take into consideration what percentage of men are rapists. If one man in every 10,000 are rapists, then you'd be spending the manpower to stop and search 10,000 men before this caught someone. And that manpower, if it were focused on more specific indicators, or causal indicators, could probably be much more effective at stopping crime, than by randomly searching every man since rapists are more likely to be men.
The same thing applies to racial profiling. Don't show me statistics about what percentage of criminals are of each ethnicity. Show me statistics on what percentage of each ethnicity are criminals. If it's not a significant percentage- which I imagine it wouldn't be- then think about how much time and manpower would have to be output statistically before you caught criminals through racial profiling, and how much better those resources could be applied when looking for directly related factors instead on coincidental factors.
I would say that you'd be right in saying that a high percentage of people who are in a gang are criminals, and so we should check people out who wear known gang colors. I'd say a high percentage of people who are lurking about on commercial property late at night, after it's closed, are likely to be criminals. I'd say that if a specific building has known drug activity, and people visit that building regularly, there'd be a high enough percentage of them that were involved, to make searching people based on that effective.
However, for racial profiling to be effective, you'd have to show me that a large enough percentage of people of that ethnicity are criminals that it would turn out significant results to do that, and I don't think those numbers are there. If they are, please correct me.
And all of that is aside from the moral implications of racial profiling, or what it costs in terms of perpetuating attitudes and resentment in the country.
The by race is comparative. And, again, the stop and frisk law, since its inception, has seen the crime rate lowered to 1960 levels. I know this is correlation, but you are saying that it may not be effective because of the low amount of criminals vs average joes out there. I'm very sure that police don't use JUST race, but also a myriad of other factors when making a stop. The discussion here is whether or not the incorporation of race into that myriad is a statistically smart move. I argue that it is, at least, an increase in effectiveness.
Post by
168916
This post was from a user who has deleted their account.
Post by
MyTie
What do you mean "increases predictive ability"?
Just "makes the predictive model guess correctly more often".
If this could be studied, which it can't, but it could, what would your hypothesis be?
Post by
ElhonnaDS
@MyTie- Which is why I don't think that they have an automatic case for racial profiling on this law, because results like those indicate they're using much more specific indicators than race. I may be wrong, but I don't think those positive results are a case for racial profiling working. I think they make a case that the cops are assessing who to stop on a number of more relevant factors.
There may be specific cops who are racially profiling, but it can't be the MO of the majority because it wouldn't be effective.
Post by
MyTie
@MyTie- Which is why I don't think that they have an automatic case for racial profiling on this law, because results like those indicate they're using much more specific indicators than race. I may be wrong, but I don't think those positive results are a case for racial profiling working. I think they make a case that the cops are assessing who to stop on a number of more relevant factors.
There may be specific cops who are racially profiling, but it can't be the MO of the majority because it wouldn't be effective.
I don't think it is an argument for racial profiling. I'm just saying that it is more effective. I get why people don't want it, but to argue that it would at least INCREASE effectiveness is beyond me.
Imagine you have a field of upside down solo cups. 60% are yellow, and 40% are red. 10% of the cups have a golf ball under them. Red cups are 7x more likely to have a golf ball under them than the yellow cups. That means for every yellow cup that has a golf ball under it, there are seven red cups with golf balls under them. You get to pick one cup, and turn it over. Your life depends on this choice. You get one chance. What color cup would you pick?
To say that other factors should be ignored, is not accurate, but to say that race is a factor that should be
ignored
, because it doesn't increase effectiveness of the search for crime is.... well... it's mathematically false.
Post by
ElhonnaDS
@MyTie- Now lets pretend that for every ball you don't find, someone dies. And while there are 200,000 cups in the field, you only have the time and manpower to check 10,000 of them.
Now lets say that the only differences between the cups aren't the color. Lets say that the upper right quadrant of the field was where people started, and we know that most probably put balls closer to their starter area. Lets say that there are a number of larger cups that hold the ball more easily, and were more likely for physical reasons to have the ball under them.
Now the reason that there are more red cups with balls under them is that there just happen to be more red cups in that part of the field, and of that shape. If you are looking for balls in that part of the field, and under cups of that shape, then those will be the most effective ways of finding the balls (since you have a limited number you can check because of time and manpower). If you tell some people to check red cups, because those happen to show up statistically more often in the categories that are causal to there being a higher chance of there being a ball, then you will have people picking up red cups when the other factors- the ones that actually increase the chance in and of themselves- aren't there, and wasting the chance to use those searches on cups that are in the right area and of the right shape. You'll also have people skipping cups that are the wrong color, but have all of the causal indicators, because you're using color as a false disqualifying factor.
The argument is that it distracts from the causal indicators, being a coincidental indicator, and that because you can only search a limited number of people, it actually reduces the number of successful criminal apprehensions in that search group by broadening the field of search unnecessarily.
Post by
MyTie
I'm not using it as a false disqualifying factor. I'm not saying ignore other factors. I'm not saying ignore yellow cups, or check every red cup regardless of other indicators. What I'm saying is that the red cups are 7x more likely to have a golf ball under them, and that is a pretty hefty statistical probability, and shouldn't be ignored. It should be included when choosing cups.
Post Reply
You are not logged in. Please
log in
to post a reply or
register
if you don't already have an account.