Author Topic: Another Case of Racial Profiling  (Read 1330 times)

LC

  • Hero Member
  • *****
  • Posts: 3457
Re: Another Case of Racial Profiling
« Reply #10 on: August 13, 2019, 08:43:12 AM »

"Why did you pull me over?" Because your race has a higher probability of committing crimes

"Why didn't you give me a loan?" Because your race has a higher probability of default

 :-\

Quote
Cops profile all the time based on their experience and intuition and check those that are most likely to be offending
...
What I see here is most of the times someone complains in the media to be only checked because of racial profiling, they were in fact caught with something illegal, making the police officer correct in his decision to pull that person over!
And their experience and intuition is most likely wrong. For example:

In a study of 20MM police stops in North Carolina (which was the first state to require this data be collected), white people had a higher probability of being caught with illegal contraband at traffic stops (36%) than black (33%) or hispanic (22%) people, yet had the lowest probability of being stopped in the first place by police.

Now, this is a complex topic. Here is an analysis you may find interesting:
https://openpolicing.stanford.edu/findings/
"Lethargy bordering on sloth remains the cornerstone of our investment style."
----------------------------------------------------------------------------------------
brk.b | irm | mo | nlsn | pm | tap | v | vz | wm


Gregmal

  • Hero Member
  • *****
  • Posts: 1905
Re: Another Case of Racial Profiling
« Reply #11 on: August 13, 2019, 09:17:06 AM »

"Why did you pull me over?" Because your race has a higher probability of committing crimes

"Why didn't you give me a loan?" Because your race has a higher probability of default

 :-\

Quote
Cops profile all the time based on their experience and intuition and check those that are most likely to be offending
...
What I see here is most of the times someone complains in the media to be only checked because of racial profiling, they were in fact caught with something illegal, making the police officer correct in his decision to pull that person over!
And their experience and intuition is most likely wrong. For example:

In a study of 20MM police stops in North Carolina (which was the first state to require this data be collected), white people had a higher probability of being caught with illegal contraband at traffic stops (36%) than black (33%) or hispanic (22%) people, yet had the lowest probability of being stopped in the first place by police.

Now, this is a complex topic. Here is an analysis you may find interesting:
https://openpolicing.stanford.edu/findings/

Would you care to bring up the statistics on % of violent crime broken down by race?

Tim Eriksen

  • Lifetime Member
  • Hero Member
  • *****
  • Posts: 766
Re: Another Case of Racial Profiling
« Reply #12 on: August 13, 2019, 05:56:45 PM »

And their experience and intuition is most likely wrong. For example:

In a study of 20MM police stops in North Carolina (which was the first state to require this data be collected), white people had a higher probability of being caught with illegal contraband at traffic stops (36%) than black (33%) or hispanic (22%) people, yet had the lowest probability of being stopped in the first place by police.

Now, this is a complex topic. Here is an analysis you may find interesting:
https://openpolicing.stanford.edu/findings/

What you cite doesn't prove your point.  In fact it partly proves the opposite.  The police stops were due to violations.  Thus the statistics show that violations are not the same by race, unless you believe that officers are intentionally seeing that the violator is white and purposely letting them go, which is possible, but you have not presented evidence for that.  The violation rate appears to be much lower for whites in the study.

It is totally separate to analyze searches and success rates on those searches.  Blacks were searched at much higher rates yet still had nearly the same rate of finding contraband.  Why?  Is that bias, or officers being smart since their success rate was consistent?  Logically bias would result in a lower success rate.  There is insufficient data to know if each race was just as likely to have contraband.   

Another point I would make is I can't find any corroboration for your statistic.  I looked at some of the data and the probability of whites being caught with contraband were lower in the two groups I looked at (Highway Patrol and Raleigh).  I didn't look at others and couldn't find a summary of all the data anywhere.     

rukawa

  • Hero Member
  • *****
  • Posts: 900
Re: Another Case of Racial Profiling
« Reply #13 on: August 13, 2019, 07:41:52 PM »
Quote
Now, this is a complex topic. Here is an analysis you may find interesting:
https://openpolicing.stanford.edu/findings/

From the site:
Quote
Becker proposed looking at search outcomes. If officers donít discriminate, he argued, they should find contraband ó like illegal drugs or weapons ó on searched minorities at the same rate as on searched whites.

Seem reasonable.

Quote
In our data, the success rate of searches (or the hit rate) is generally lower for Hispanic drivers compared to whites; so the outcome test indicates Hispanics face discrimination. For black drivers, search hit rates are typically in line with those of white drivers, indicating an absence of discrimination.

So they basically found no bias against black. So what did they do?

Quote
Beckerís outcome test is a compelling measure of discrimination. But itís also an imperfect barometer of bias....To address the shortcomings of the outcome test, we built on Beckerís ideas to develop a more robust statistical measure of discrimination: the threshold test

They invented a more complicated test and got the answer they were searching for in the first place. To me this study demonstrates more about the bias of the researchers than it does about racial bias. A researcher can keep doing this forever. They can slice and dice the data, reject portions of it, conduct increasingly complicated tests and basically find an answer that they like.

I regard all these statistical analysis as basically complete junk and I don't really think anyone...even professional statisticians like this guy:
https://statmodeling.stat.columbia.edu/
have any clue as to how to make statistical arguments reliably or intelligently. The real thing people are trying to do is derive casual arguments from statistics. But statistics are just associations. To make a casual argument requires a casual model and then subsequently a careful testing of the casual model.