[Peace-discuss] Fw: details on numbers I cited in the Urbana city council meeting last night

[Randall Cotton]

For those who would like to know more about the Illinois racial profiling study
results for this year....
R

----- Original Message ----- 
From: "Randall Cotton" <recotton at earthlink.net>
To: "Mike Monson" <mmonson at news-gazette.com>; "Charlie Smyth"
<csmyth at sbcglobal.net>
Cc: "William Brown" <urbanabill at yahoo.com>; <bowersox at prairienet.org>;
<hstevenson4 at yahoo.com>; <robertlewis10 at insightbb.com>;
<lynnecbarnes at hotmail.com>; <droberts at uillinois.edu>; <chyn at ojctech.com>;
"Laurel Prussing" <llprussing at city.urbana.il.us>; <bkwalden at city.urbana.il.us>
Sent: Monday, August 28, 2006 1:10 AM
Subject: Re: details on numbers I cited in the Urbana city council meeting last
night


: I've done some more analysis of the IDOT report regarding black vs. white
: traffic stops.
:
: I mentioned at the last Urbana City Council meeting that the numbers indicate
: blacks are over three times (3.26 times) as likely to be stopped as whites in
: Urbana. I mentioned that this statistic is worse than Champaign (2.79). I've
: since checked all the larger surrounding Illinois cities. Urbana ranks at the
: very bottom of the list:
:
: Danville: 1.84
: Bloomington: 2.48
: Peoria: 2.56
: Champaign: 2.79
: Normal: 2.92
: Springfield: 2.98
: Decatur: 3.02
: Urbana: 3.26
:
: The average for the state is 1.26 .
:
: This number is just the rate of black stops divided by the rate of white stops
: (by driving population). It's a very simple calculation that can be made just
: with IDOT data (and in particular for Urbana, from the numbers in tables 1 and
2
: on page 7 of Chief Bily's report):
:
: black stops/black drivers
: ---------------------------
: white stops/white drivers
:
: which is:
:
: 1005/3875
: ------------ = 3.26
: 1707/21428
:
: It seems to me that this is a strong indicator of a possible racial profiling
: problem and should attract urgent closer scrutiny. The driving population
: numbers are estimates, of course, (it's the 2000 census data for people age 15
: or over), but regardless, the magnitude of this number is so overwhelming and
: compares so poorly with the surrounding major cities and the state average
that
: (as Charlie says) it hits you right between the eyes.
:
: Note that this number is completely unaffected by whatever the Asian
population
: happens to be. We could have an Asian population of 1% or 90%, but it would
: still be the case that in Urbana, black drivers are still being stopped at
over
: 3 times the rate that white drivers are.
:
: You don't have to take my word for it, just check the attached Excel
: spreadsheet. All the numbers come directly from the IDOT reporting available
: on-line at:
:
: http://www.dot.state.il.us/trafficstop/racialprofiling.html
:
: Specifically, from these three documents:
:
: http://www.dot.state.il.us/trafficstop/2005summary.pdf
: http://www.dot.state.il.us/trafficstop/2005study.pdf
: http://www.dot.state.il.us/trafficstop/2004summary.pdf
:
: R
:
: ----- Original Message ----- 
: From: "Charlie Smyth" <csmyth at sbcglobal.net>
: To: "Randall Cotton" <recotton at earthlink.net>; "Mike Monson"
: <mmonson at news-gazette.com>
: Cc: "William Brown" <urbanabill at yahoo.com>; <bowersox at prairienet.org>;
: <hstevenson4 at yahoo.com>; <robertlewis10 at insightbb.com>;
: <lynnecbarnes at hotmail.com>; <droberts at uillinois.edu>; <chyn at ojctech.com>;
: "Laurel Prussing" <llprussing at city.urbana.il.us>; <bkwalden at city.urbana.il.us>
: Sent: Tuesday, August 22, 2006 4:35 PM
: Subject: Re: details on numbers I cited in the Urbana city council meeting
last
: night
:
:
: : Randy,
: :
: : I'm not disagreeing with your analysis. I said right at the beginning
: : that the numbers show that 1/3rd of the traffic stops are to African
: : Americans and only 8% are Asian.  Should we be outraged, maybe, maybe
: : not. Is there profiling, maybe, maybe not. You can't prove it
: : statistically one way or the other from the data presented. But you
: : can ask why do we see the numbers we see. One needs to lay out the
: : hypotheses and try to figure out how to test them. Having been the
: : victim of profiling when I was young, I'm not insensitive to this
: : issue (I used to travel a lot when I was 15-25 by air when hijacking
: : planes to Cuba was the terrorism of the day or being suspected of
: : smuggling drugs from Ecuador where my family is from -- I used to
: : have shoulder length hair along with a very full beard etc etc)
: :
: : First, none of the data presented have any kind of statistical
: : sampling error associated with them. These should have something,
: : they are based on estimates.  I need to read the NW study but the
: : numbers can't possibly be "exact" and so given the statistical
: : uncertainty of the whole exercise I much prefer to think in
: : approximates here because whether it's 2.3 or 2.5 or 2.7, there's
: : really no way to tell what the "real" numbers possibly are and hence,
: : trying to come up with some kind of "exact" number is really ludicrous.
: :
: : So, if you start digging a bit you see a number of interesting
: : things: look at table 4 where accident statistics are reported.
: : Caucasians are higher than would be expected, Asians lower, African
: : Americans and Hispanics about what you'd expect given uniform
: : distributions and all that.  From this accident data do we conclude
: : that Caucasians are lousy drivers and Asians really good? That's just
: : as valid as saying that there's racial profiling going on. What we
: : really need to be asking in all this is why are we seeing the numbers
: : that we see, not just in one particular table, and what additional
: : analyses are needed. It really is important to look at all of the
: : numbers presented, not just one particular measure.
: :
: : So, one can hypothesize that the Asian community, being mostly
: : students that live in and around campus, don't drive as much and when
: : they do, they stay close to campus (note that in the break down by
: : beat, the only time Asian numbers are anywhere near where you would
: : expect is in the campus area). (I wrote this last night before going
: : to bed and it echos what Willam Brown stated in his email).  Then you
: : have the whole complication of half of the campus beat being covered
: : by the UofI police department. How do you figure that huge, mostly
: : Caucasian/Asian population into the figures. If you were to remove
: : some portion of Caucasian and Asian population from the pool of
: : "Urbana" drivers, probably about 10,000 people, then the African
: : American population over the rest of the community goes up.  In the
: : accident situation is it that Caucasians drive more than average (and
: : there are likely economic factors here that come into play) and if
: : accidents correlate to miles driven, Caucasians are going to have
: : more accidents? Do a majority of Asians in our community drive less
: : and when they do, do they go south on Race to Windsor and drive to
: : the Savoy Walmart? Or are they traveling straight to Meijer and north
: : Prospect via Florida/Kirby (ie. putting them mostly in Savoy or
: : Champaign for their driving miles). Another aspect of this is bus
: : ridership and the racial breakdown there. How does this skew the
: : numbers or is it so small, and again, messed up by students and the 3
: : community situation that it has no effect?
: :
: : Another interesting table shows that 25% of the speeding tickets go
: : to African Americans. Again, a very small number to Asians. So, look
: : at where STEP patrols are going on with over 40% in beat 6-1 and from
: : talking to the police department it's mostly right along Lincoln Ave.
: : Lowest in Beat 6-3, and the 2nd highest is in 6-4 which is twice as
: : big and from talking to the officers, I get the impression that it's
: : more uniform though concentrated around Windsor. Bottom line here is
: : that Beat 6-1 has a very large black community, 6-4 has the next
: : highest black community. If your policing is concentrated in a
: : particular area, your stops are going to reflect the population
: : numbers of that area. Again, another compounding factor: half of beat
: : 6-3 is university, covered by another police department, and much of
: : the population doesn't have or use cars regularly. If 1000 of the
: : tickets issued are from STEP in 2 particular parts of town, the data
: : is skewed. Another interesting statistic that wasn't even discussed
: : last night is the racial breakdown of calls for service, and I think
: : 25% are African American. This is why we need GIS and census tract
: : info to overlay on all of this, we really don't know the underlying
: : population structure.
: :
: : If you want to accuse the police of something, maybe it's making
: : money from speed traps on Lincoln Avenue! I know that if I get
: : nailed, it's going to be coming off of I74. I was looking for the
: : concept of speed habituation last night!
: :
: : Then there are all kinds of cultural based hypotheses that need to be
: : tested such as do black drivers speed more than other groups? The
: : numbers already show that 60% of the speeding tickets are in the
: : 17-31 age group so we know youth is an association but we also need
: : to know the breakdown of this group - maybe this group is 25% African
: : American? (I don't believe this is the case but I'm trying to make
: : the point that we really do need better base numbers to work from -
: : we do know that age distributions for different groups is different
: : but how different I don't know. It will be interesting to follow the
: : Hispanic numbers over time if in fact this community is growing in
: : our area). What's the gender breakdown?  What about loud radio
: : systems - is there a cultural link?  And AC Connelly was right, loud
: : radios goes in the equipment category.
: :
: : Socio-Economics? (this one strikes me as the easiest one to test
: : quickly by simply correlating age of vehicle with race of driver. If
: : you were to find that a disproportionate number of one racial group
: : was stopped after adjusting for age/condition of vehicle then you
: : might have something. Eg. If you drive an old beater with a tail
: : light out, you're more likely to get stopped versus if you drive an
: : old beater with a tail light out and you are black, you will get
: : stopped. I wonder if there are other creative ideas and analyses
: : along these lines that would get us a better picture.
: :
: : Bottom line is that I would be much happier with a real statistical
: : approach to this whole thing but then, it would probably be extremely
: : expensive if even doable. Lots of assumptions and lots of agendas -
: : and this doesn't make for good conclusions.
: :
: : Yes, profiling happens in our society but I hope it's becoming more
: : of thing of the past (though one has to wonder how it is to be an
: : Arabic male in the US today), particularly in our community, and that
: : we can move forward. What we really need is to have as many tools at
: : our disposal to oversee the police department and others in authority
: : so that no one abuses their power. Remember too that this council is
: : working diligently on a Civilian Police Review Board. Similarly, we
: : need to address socio-economic and very related, educational
: : opportunities, in our communities.  We need to address underlying
: : racial stereotypes and develop an understanding of cultural
: : differences. These are long term issues that others have not solved
: : but we can keep trying!
: :
: : -charlie
: :
: : At 11:39 PM 8/21/2006, Randall Cotton wrote:
: : >First of all - one quick note - if I'm quoted or identified in a news story
: on
: : >this, I'd prefer to be identified as a member of Champaign-Urbana Citizens
: for
: : >Peace and Justice. Though I'm also a member of AWARE, CUCPJ is the context
in
: : >which I'm doing this activist work. All four folks who spoke on this
: : >issue last
: : >night are members of CUCPJ.
: : >
: : >As I indicated, IDOT's (minority) "disparity index" is calculated by:
: : >
: : >% of minority traffic stops in Urbana
: : >-------------------------------------------
: : >% of minority driving population in Urbana
: : >
: : >which is:
: : >
: : >44.12%
: : >-------
: : >30.62%
: : >
: : >( these numbers are on page 1 of the IDOT report for Urbana:)
: : >
: : >and this comes out to the ubiquitous 1.44 figure
: : >
: : >But this masks the severity of the problem for the *black* minority. If you
: do
: : >the same calculation, but just for blacks, you get:
: : >
: : >% of *black* traffic stops in Urbana
: : >-------------------------------------------
: : >% of *black* driving population in Urbana
: : >
: : >which is:
: : >
: : >32.90%
: : >-------  =  2.7
: : >12.19%
: : >
: : >Again, I specifically cited the 2.7 *black* "disparity index" in my
: : >comments to
: : >the council, and this number is *far* worse than the 1.44 number cited by
: IDOT
: : >for *minority* disparity index. By lumping all minorities together and not
: : >breaking it out by race, the severity of the problem in the black community
i
: s
: : >not revealed. This number is worse in Urbana than in Champaign (Champaign's
: : >*black* disparity index is 2.29).
: : >
: : >In case you're not clear on this, the 32.90% number can be calculated from
: : >numbers in page 1 of the IDOT report:
: : >
: : >specifically:
: : >
: : >% of *black* traffic stops in Urbana =
: : >
: : >black traffic stops
: : >------------------  X 100
: : >total traffic stops
: : >
: : >which is:
: : >
: : >1005
: : >-------  X 100
: : >3055
: : >
: : >and that comes out to:
: : >
: : >32.90%
: : >
: : >The 12.19% number comes from the chief's report (top right hand corner of
: page
: : >7). Mike has reported that number as "black drivers represent just
: : >12 percent of
: : >the city's driving population" in his July 23, 2005 article last year.
: : >
: : >Now, I subsequently went on to describe *another* way that IDOT's
"disparity
: : >index" under-represents the severity of the problem. As you know,
: : >the so called
: : >"disparity index" is a calculation of how much more likely one is to
: : >be stopped
: : >by police in Urbana than your racial driving population percentage would
: : >indicate (and you can crunch a number for minorities as a whole, which is
: what
: : >IDOT does, or specific minorities, like I did above for blacks).
: : >
: : >But the explanation of this figure is convoluted and confusing, especially
: the
: : >part "than your racial driving population percentage would indicate", and I
: : >think it's lost on just about everyone. I think that folks generally read
: that
: : >1.44 number and think "oh, minorities (maybe they even think *blacks*) are
: 44%
: : >more likely to be stopped than whites". And if they do make that
conclusion,
: : >then they are sorely deceived. The number they're imagining,  a much more
: : >tangible "real world" comparison of how much more likely a black
: : >driver will be
: : >pulled over than a white driver, is actually *immensely* higher than 44%.
: : >
: : >Here's the calculation for that:
: : >
: : >You start with:
: : >
: : >% of *black* traffic stops in Urbana
: : >---------------------------------------
: : >% of *white* traffic stops in Urbana
: : >
: : >which would be all you need if the black and white populations were equal.
: : >
: : >But since they aren't, you have to multiply by a weighting factor,
: : >specifically:
: : >
: : >% of white driving population in Urbana
: : >-----------------------------------------
: : >% of black driving population in Urbana
: : >
: : >For example, if the white driving population were twice as big as the black
: : >driving population, the weighting factor would be 2. It might help to think
: of
: : >it this way: going along with this example, since there are half as
: : >many blacks
: : >as whites in the community, then if magically the black driving population
: : >doubled, you'd have an equal black/white population now, but the
: : >number of black
: : >traffic stops would double too. That's why you need to multiply by two to
: : >accurately represent black vs white traffic stops.
: : >
: : >Getting back to the real world numbers, all together, you get:
: : >
: : >32.90%     69.38%
: : >------  X  ------ = 3.35
: : >55.88%     12.19%
: : >
: : >That is, if you're driving down the street in Urbana, you are 3.35 times
more
: : >likely to be stopped by Urbana police if you're black than if you're white.
: : >Another way of saying this (and what I said in my comments to the
: : >City Council)
: : >is that you are 235% more likely to be stopped by Urbana police if
: : >you're black
: : >than if you're white.
: : >
: : >The number for Champaign this year is also bad, but not nearly as bad as
: : >Urbana's (2.88 vs 3.35).
: : >
: : >BTW, the 55.88% number comes from the top of page 1 of IDOT's
: : >report. The 69.38%
: : >number for % of white driving population is calculated the same way as for
: : >blacks (which I described in detail above).
: : >
: : >I've attached the spreadsheet I used to figure all this out.
: : >
: : >Charlie, you mentioned the odd numbers for Asians vs. Blacks. You'll be
: : >interested to know that blacks are 5.34 times more likely to be stopped
than
: : >Asians in Urbana. That's in the spreadsheet.
: : >
: : >R
: :
: : Charlie Smyth                   csmyth at sbcglobal.net
: : 805 S. Coler                    217-367-2813
: : Urbana, IL   61801              217-333-9477 (W)
:
: ------------------------------------------------------------------------------
: -----------------------------------
: : Bill Gates is the world's most spammed person -- he receives 4 million
emails
: : per day, most of it spam. It's kind of like that scene where Dr.
: : Frankenstein's
: : monster killed him. (DGB)
: :
:
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 2005 Traffic Stop Analysis.xls
Type: application/vnd.ms-excel
Size: 27648 bytes
Desc: not available
Url : http://lists.chambana.net/cgi-bin/private/peace-discuss/attachments/20060828/8ec6904f/2005TrafficStopAnalysis-0001.xls