HomeMy WebLinkAbout2014-06-16 TranscriptionJune 16, 2014 Iowa City City Council Special Work Session Page I
Council Present: Botchway, Dickens, Dobyns, Hayek, Mims, Payne, Throgmorton
Staff Present: Markus, Fruin, Dilkes, Karr, Hargadine, Steffen, Hart, Gaarde, Schwindt,
Bowers
Others Present: McCarthy (UISG)
Police_ Department Traffic Stop Demographic Analysis (IP3 of 6/12 Info Packet) :
Hayek/ (starts midsentence) ... to, uh, this work session. We have one agenda item this evening,
which is a presentation of the Police Department traffic stop demographic analysis.
Chief!
Hargadine/ Uh, good evening! Sam Hargadine, Chief of Police, in case we're recording! (noises
on mic) The Iowa City Police Department has been collecting demographic data on
traffic stops since 1999. In January of 2004, a report titled "Traffic Stop Practices of the
Iowa City Police Department, From January 1 Through December 31, 2002," was
published. The research team was from the University of Louisville and the report is fre
...frequently referred to as the "Louisville Study." In 2006, the command staff was
approached by Dr. Christopher Barnum. Dr. Barnum was interested in obtaining the data
from the Iowa City Police Department traffic stops to complete a study for a research
project that he was involved in. Dr. Barnum agreed to use a synonym for Iowa City if his
paper were to be published. Originally, Dr. Barnum completed six months worth of data,
and we met with Dr. Barnum after he completed the first review and determined that a
multi -year study was necessary to really understand and compare traffic stop patterns.
Additional raw data was provided to Dr. Barnum for the years of, uh, 2005, 6, 7, 10, 11,
and 2012. ' The datas ... the data from years 2008 and 9 were found to be corrupt and
unusable. Dr. Barnum is an Associate Professor of Sociology and Criminal Justice. He's
the Director of Graduate Studies, Masters in Criminal Justice at St. Ambrose University.
Dr. Barnum has published in numerous peer review journals, peer reviewed
jou ... journals, book chapters, and technical reports. At this point I'd like to turn ... turn it
over. (noises on mic)
Barnum/ My name is Chris Barnum. I'm from St. Ambrose University. This is, uh, Bob Perfetti
who's a graduate assistant who's helped me with this paper, and uh, over ... back over
here is Shane Erickson who's also a graduate, uh, student who's going to be assisting in
the future on this project. Um, the Chief did a nice job of, uh, sort of setting the
background for ... the, uh, the processes that took place prior to ... to looking at the data.
Um, I have quite a few slides on this PowerPoint, so I think maybe the best way to go, as
I give the PowerPoint, is feel free to interrupt me and ... and ask questions, uh, slide by
slide because, uh, it comes at you ... there's quite a lot of information that comes at you.
Um, and so ... as the Chief said, we look ... we're looking at data from 2005 through 2, uh,
2012, with the exception of 2008 and 9. The data were missing, uh, I believe that the
Department had some issues with data being overwritten. They were switching software
at that time and so, um, those data are missing, which is kind of an interesting, uh,
phenomenon because 2008 and 2009 are going to be kind of a pivotal time period in this
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 2
study as we go forward... we're going to look at. Uh, the way we're going to be looking
at the data is we're going to be comparing the older data from 2005/2007 to the more
recent data, 2010 through 2012. And, uh, as the study progresses you'll see that there is
some changes in the data, uh, in that time period, and the changes are ... are, um,
probably, and I'll explain this more as we go forward. Probably related to, uh, an
in ... increase in crime that occurred on the southeast side of Iowa City during that time,
which, uh, caused the Police Department to change some of their tactics and patrol
procedures, and uh, that resulted in an increase in disproportionality. So ... when doing a
study like this, there is essentially, uh, two major outcomes that we look at. And the first
one is whether or not a stop is made, and the second one is the outcome of the stop. So
the first thing, the first half of the talk is going to be about the percentages of stops that
were made in terms of demographics, the race of the driver versus a base line. So we're
going to be comparing Iowa City Police data to a base line, and the base line was
established from, uh, essentially, um ... roadside observations and Census data. And then
after we talk about that, we're going to talk about the outcome of a stop. So the outcome
of a stop is what happens to the driver after the stop is over. You can receive a warning.
You can receive a ticket. You can be arrested, uh, all kinds of things like that. So we're
going to look and see if there's any differences in terms of, uh, racial demographics, in
terms of outcomes and stops. Uh, this slide that I have on here now, uh, and I'll talk
more about this as we get into the talk a little bit, this just shows that, uh ... just to sort of
set the picture, um ... what our observers saw when they were out watching traffic. So
starting in the year 2007 and all the way through 2012. We had people on the streets of
Iowa City, in various, uh, spots, watching traffic and essentially what they would do is
just record the race and gender of the drivers that they saw. And race is broken out as
white, black, Asian, or other, and then gender was male or female. And as they watched,
uh, they would simply check a box, at various locations. We did this on the ... Monday
through Sunday, seven days a week, and we did it all timeframes — in the morning,
evening, late night, and the ... the process of observing, uh, the initial wave took place,
um, in the summer and fall of 2007. Then we updated in 2010 and I believe 2012 we did
some more... observing. And as you... as you look at that you can see we, uh, the
initial ... this is the data for the initial wave. We made 22,000 observations, um ... uh, the
percentage of white drivers that we saw was 88 %. See if my laser works here ... yeah,
right there, and then, uh ... our observers saw about 3.83% of the drivers on the streets in
Iowa City were black. Asian were 3.88. Uh, some that we classified as other were about
4 %, and then over in this column right here you can compare it to what the 2010 Census
data said. So, uh, the Census said there's about 83% white, 5.8% uh, black, 6.9 — roughly
7% Asian. So you can see there's a pretty close correspondence to ... to what our
observers saw and what the Census data, which gave us some confidence that our
baseline's probably accurate, and as we go forward, I'm going to show you that, uh, we
did some oversampling, particularly on the southeast side, and um... so we have a ... I... I
think you'll agree with me when we get to that point that the ... the baseline is a pretty
good measure to compare against. Um, the thing with the baseline that you want to think
about as we're moving forward is ... the baseline needs to remain stable. So in other
words the baseline in 2005 needs to be the same as the baseline in 2012. So we spent a
lot of time making sure that that baseline wasn't changing, cause if the baseline changes,
that would increase, uh, levels of, uh, disproportionality. But it would be because the
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 3
population was changing, not the tactics of the Police Department. So we'll talk a lot
about that as we move forward too. Okay! So, this first box of the percentages of drivers
stopped broken out by race for the year 2005 by the Iowa City Police Department.
And ... so they ... they made about 10,000 stops. These are stops for which all information
is available to us. Sometimes there was data missing and so we couldn't include those in
there, and um ... when there is missing data it'll... there'll be a column that'll show that
up. So you can see, uh, in 1985...84% of the stops were whites, about 9% black, 3%
Hispanic, and 2% Asian, and then so on down the line. Which interestingly enough are,
um, very close to the numbers of the 2001, uh, Louisville study found. They found very
similar numbers to that. So ... in 2001 we know that the numbers were very similar to
what they were in 2005, and that's going to be kind of important when we move forward.
So these are the police beats that, uh ... Iowa City uses.
Payne/ I have a question.
Barnum/ Yes!
Payne/ As far as when the officer enters the race into their computer to keep track of this, they're
just guessing the race, correct? They're not asking the person — they're guessing.
Barnum/ Yeah, I don't believe they ask. I think it's their perception of race.
Payne/ Okay. Thank you!
Barnum/ So there's four police beats in Iowa City. Now you're going to see in the data as we
move forward when we break this out by beat we're going to use five beats. We're going
to use those four beats and then beat number five are people who were not assigned to a
beat, and most of those officers were either canine officers or members of the SCAT unit,
the street crime, uh, action team unit. So they ... they're not assigned to a beat, but they're
making traffic stops, so we call those officers `beat 5' as we move forward when you see
that. Okay? So here it is broken out by beat for 2005. Bottom ... row, right here, shows
the percentages on each beat of minority members that were stopped. Now we classify,
uh, minority members as anybody who is not white or Asian. And the reason we group
white and Asians together is because in the research, uh, nationally, and it was the case
here in Iowa City as well, Asians tend to be stopped at lower numbers than their baseline,
uh, percentages in the population. So they're under - represented. So if Asians were
grouped with, uh, African Americans, for example, that would tend to depress the
number of minorities that were stopped. So you group Asians and whites together, and
that is the clearest way to identify disproportionality, and that's not a problem doing that
as long as you're consistent. So we do that for each year of the study, from 2005 all the
way through 2012.
Throgmorton/ Chris, if I could follow up on that. I'm ... I'm over here!
Barnum/ Oh!
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 4
Throgmorton/ Yeah, if I could follow up on that.
Barnum/ Yes!
Throgmorton/ Uh, my sense was that there... your... your report uses the term `minority' in a
rather flexible way in the sense that sometimes it refers to African Americans and others,
and other times it refers to African Americans. Uh, so uh ... I'm wondering if you think
that might have any implications for what you find in the study.
Barnum/ Yeah, I ... uh, I agree with your point, particularly when we look at the Census data on
the beats. We're looking at African Americans, uh, rather than, urn ... grouping them all
together, and the problem ... the problem with the Census data is that the Hispanic
classification isn't clear on which group they go with. So when we look at the Census
data, actually we have to, uh, talk about one group in particular, and since African
Americans are the largest, that's the one that we used. But I think it's clear from context,
uh, and if...if it's not at a ... at an individual point, just ask me, uh, what `minority' is
referring to. Okay? Now the thing to look at on these graphs as I go through them is, uh,
beat 2, and beat 5, for the most part. See, you can see like we just saw about 14% of the
stops are minority members in the year 2005. We set our baselines 10 %. So, uh, 10%
is ... are minority members, um, 90% non - minority members, according to our baseline
from the ... from our observations. So there's about 4% over, but that doesn't necessarily
mean anything. That's a small amount over. Our observations are estimates. There can
be several points, up or down, uh ... around the mean of what we saw. So it's just kind of
a guess. It's a ... it's an estimate. So 4% isn't ... isn't anything that's particularly alarming
or high. Uh, and as you look through each of these, they're all fairly low. And now
we're into 2006. These are just ... it's just broken out the same way. If you look at this,
there's not much change here. So we'll go to it by beat. And ... we look at 2006 by beat,
uh, this is 17% on two, five is 20 %. So they creeped up a little bit and the overall
percentage is 15 %, which creeped up a little bit, but that again can be noise. That can be
due to just, uh, random fluctuation. I'm just going to go through these a little bit here.
2007, there are the stops that they made. And then by beat again. So you see, uh...beat 2
is down to 13 %, beat 5 is 19 %, and the overall is 14 %. And so there really wasn't much
change. There really wasn't too much of a change right there, uh, between the years
2005 and 2007. The numbers were pretty consistent, and they were also very consistent
with what Louisville found. They classified their data a little bit differently than, uh, we
did, but ... the percentages were ... were very close. So we can be confident that from 2001
up until 2007, that the percentages being stopped by the Police Department were pretty
much the same, year to year to year to year. Okay, so then 2008 and 2009 happened, and
in 2008 there was the flood, and then in 2009 there was an increase in crime, and there's
some pretty high - profile events that happened in 2009. There was, uh, what is known as
the `Mother Day Riots, ` which got a lot of press coverage. If you Google that you can
find that in all kinds of newspapers, not just local ones. Um, and there was, um, there
was a murder on the, uh, southeast side in the Broadway area, and some armed robberies
and some, uh, high - profile events, and so at that point, the Police ... as I'll show, instituted
some changes to their pro ... uh, patrol procedures on the southeast side, which are going
to lead to an increase in disproportionality, especially among beat 2 officers — the officers
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 5
who work on the southeast side, and beat 5 officers — the officers who are not assigned
but are the special, uh, SCAT officers who can go wherever they want. Presumably
they're going to where most activity in the city is. So let's take a look at ... 2010. Uh, so
do you remember, uh, if you look at this number right here, it was in the 80s before. This
is percentage of whites. So now you can see that that's decreased. It's ... it's dropped
down into the 70s, and the percentage of blacks stopped has bumped up now into, uh,
double digits. It was around 9 before. And so you can see something's happening. The
data are starting to change. And when we look at it by beat, um, so look at beat 2 here.
You can see now 26% of the stops on beat 2 were minority members. 22% on, uh, beat 5
were minority members, and the overall for the Department is 19 %. It was 14% before.
So let me just go back real quick to ... uh, to 2007. You can see on beat 2 in 2007, the last
year for which we had data, 13% of the stops on beat 2 were minority members.
(mumbled) up ahead here to 2010. It's up to ... 26 %. So that's a pretty big increase, uh,
and so ... that's when you look at the data more closely, and you want to look over ... look
at it over time to see if this is a trend that persists. And so we'll take a look here and see
what happens.
Botchway/ One ... one second, I had a quick question.
Barnum/ Yes!
Botchway/ And so ... number of overall stops kept increasing throughout the years, correct? It
seems like ... or maybe they just fluctuated back and forth (both talking)
Barnum/ They ... the total number of stops...
Botchway/ ...total number of stops.
Barnum/ Yeah, they ... they do kind of fluctuate back and forth. There's, uh, if you wanted to say
in general there was probably a few more stops made in the second half, compared to the
first half, but it's ... it's not a dramatic difference. Now 2007 was not a full year's worth
of data. That data started crashing in the end of November, so... so we don't have a full
year for that, so that explains the lower numbers in 2007.
Botchway/ Okay, now going to 2010, where you see the 77% which you showed, and the 13 %,
um (both talking)
Barnum/ Yeah.
Botchway/ I guess I just want to be clear ... so that doesn't mean, you know, there was a drop
necessarily in crime from the 83% that was shown earlier. It was just that the focus was
now, you know, on a particular sector of the community that would rise to more
disportionate... (both talking)
Barnum/ That's correct. These are just traffic stops. Uh, we ... this data don't show anything
about crime. They ... uh, so we'll look at the crime statistics here in a minute, but yeah,
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 6
this is just the percentages based on different categories of people who were stopped by
the police. And again then you would compare that number to our baseline. So the ... the,
kind of the ... just the rough number to keep in your ... in your mind is that 10% is the
baseline. So before we were at 4% over, and now we're up to 9% over.
Botchway/ Yeah, and I understand (mumbled) talk about crime. I was talking more about the
drop from 6 %, from 83% to 77 %, and that meaning less traffic stops...
Barnum/ Of whites.
Botchway/ Right.
Barnum/ Yes.
Botchway/ And ... I'm just I guess making the assumption we're saying that's because of the
focus towards particular (mumbled) areas.
Barnum/ Yeah. So the white percentage has to drop ... when another percentage goes up.
Botchway/ Okay!
Barnum / And so the, uh, the African American percentage went up and that ate into, if you will
(both talking)
Botchway/ ...assume; I just wanted to make sure!
Barnum/ Yep, yep, that's right!
Throgmorton/ Chris, I'd like to follow up on that a little bit. Uh ... and I'm looking at the report
itself, not so much the charts.
Barnum/ Yes!
Throgmorton/ On page 13 the report states that 92% of the drivers were white or Asian, and 8%
were `minority members.' (both talking)
Barnum/ That's the baseline ... (both talking)
Throgmorton/ ...but it subsequently uses 90/10.
Barnum/ Yeah, that was the baseline you were talking about...
Throgmorton/ Yeah, yeah.
Barnum/ Yes.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 7
Throgmorton/ Uh, so this ... I think that has meaningful consequences for how one might
interpret, uh, the ... the resulting information about stops, and here's what I mean. Uh, on
page 21 the report states, `Given a 10% minority baseline, this suggests that in 2010 to
2012 overall levels of disproportionality increased from roughly 4% to about 8 %,' which
is pretty much what you just said, but if...oh, I just lost my thing. `But if 8% was used as
a baseline, the data could be said to indicate that disproportionate contact more than
doubled from the baseline period to 2010 and 2012.' So the... different way of
interpreting, and I want to see if you would agree that one could interpret it that way.
Barnum/ It could be, but you... you're putting a lot of faith in that baseline.
Throgmorton/ Uh huh.
Barnum/ The...the baseline, I just want to emphasize, you know, it ... it's ... an estimate. It's a
sample. It's a very small slice (both talking)
Throgmorton/ Right, and I totally understand (both talking)
Barnum/ Yeah!
Throgmorton/ (both talking) ... observing and (mumbled)
Barnum/ Yeah, and ... and so ... I mean, it's going to have a margin of error up or down, but yes,
you could interpret it that way. So, um, if it were 8 %, so it would be...it would have
gone from ... 6 % ... right? In the, before ... it was 4% in the early years, so if you add 2 to
that it would have been 6% and then it would have gone from 6% to 10 % ... based on an
8% baseline, right?
Throgmorton/ I'm not ... I'm not sure what 4 and 6% you're referring to.
Barnum/ Well you ... you, as you mentioned ... as you mentioned, so it ... let's say in 2007, 14% of
the stops were minority members.
Throgmorton/ Uh huh.
Barnum/ The baseline's 10 %.
Throgmorton/ Right.
Barnum/ So 4% (both talking)
Throgmorton/ So that's what you meant by 4.
Barnum/ Yes.
Throgmorton/ Yeah.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 8
Barnum/ If that were 8% like you said, instead of 10 %, the baseline, that's 6 %. So the increase
would have been from 6% to 10 %. Right?
Throgmorton/ Uh, I guess I'm seeing different numbers. I ... I was seeing, uh, going from 8% to,
um ... whatever the number was in 2010 to 2012, which was something like 19 %. I don't
remember exactly (both talking) Yeah, so that's like going from 8 to 19, that's more than
doubled. That's the way I was...
Payne/ But ... but you're also looking at just one group compared to the minorities as a total. So
you can't ... you can't mix apples and oranges either. You can't just say African
Americans is 8, but the total is 19. That's comparing apples to oranges.
Throgmorton/ Well this is one reason why I raise ... I raise that point about the distinction, uh,
having to do with minorities. What one means by that. So ... some clarity's needed, I
think...
Barnum/ Well no it's ... the baseline is minorities, and that 19 %'s not African Americans. It's
minorities. So it's ... it's talking about minority to minority. The ... the quibble that we're
having, it's not really a quibble, but the ... the discussion that we're having back and forth
is whether that baseline's 8% or 10 %, and you know ... it's somewhere between 8% and
12 %, I would guess, but I, you know...
Throgmorton/ Okay.
Barnum/ So that's... that's, you just have to look at it that way. It's ... it's something that you
should look at more as just ... to use to judge over time. The more important question is,
is that baseline changing over time. Because if the baseline is changing, for example
if... if the number of African Americans living in Iowa City increased between 2005 and
2012, we would expect the percentage of African Americans being stopped to increase.
So that's the real question to look at, and we're going to spend a lot of time looking at
that. That's... that's... that's really the issue. That's... that's what's important to look at.
Payne/ I have a question also.
Barnum/ Yes!
Payne/ Um, and you may not know the answer to this, but I'm curious as to why the command
staff personnel, number of stops, increased so hugely over the time period also, and you
don't have to answer that right now. It's just a question that I have since we're going
through the charts that indicates that.
Barnum/ Yeah, I...
Throgmorton/ Yeah, I ... I had the same question, Michelle, and there's one year in particular
when the number of command stops is like 1,800 or something like that.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 9
Payne/ Uh huh.
Barnum/ Uh, one thing that could account for that, I don't know that this is the case, but
someone who was ... promoted half -way through the year might have been classified as a
command personnel and the stops that they made while they were still on the street
carried over with them. Yes, that's possible. And that's probably what did happen.
Okay, so we're up to 2011 now. And the percentage is 18 %. So again, it's very close.
This number is 76 %, beat 2's at 25 %. So beat 2 ... beat 2 is where the, you know, a lot of
the action is at, it looks like, where the ... where the ... increase in disproportionality is
being driven on beat 2. Now, let me just say a word, which I'm a little remiss I didn't say
it right up front and I should have. The word disproportionality simply means difference
between our baseline and ... what the police are stopping. It doesn't any way imply bias
or disparity. That's... that's a judgment call. All I'm saying is there's a difference, okay?
So that's what that term disproportionality means. So we have an increase in
disproportionality and most of it looks like it's being driven on beat 2. So...
Payne/ And isn't there a place in here also that says that ... at least part of beat 2 is 40% minority?
Barnum/ Yeah, I'm going to talk about that here.
Payne/ Okay.
Barnum/ Yeah, we'll get to that. Yep! So it...it matters where the stops are being made on beat
2 very, very much. Yeah. So we're up to our last year, 2012, and again, this number is,
uh, creeping down a little bit. These numbers are fairly stable. So here we are on beat
again. 25% again on beat 2. Beat 5 is up to 29 %. Those are the special officers, the
special enforcement officers, and the overall level of disproportionality with rounding
errors is about 18 %. So ... this next slide shows you beat 2 over time, what happened.
And you can see it pretty clearly. The dividing line of 2008 and 2009, there's... there's,
uh, when you look at it, there's a difference between the later years and the early years.
Uh, that's not in your report. This is just for the PowerPoint. So, um, and the reason I
didn't put bar charts in the report is because you'll notice when we go through these bar
charts, I'm going to do it for beats, uh, beat 5 and beat 3, as well, on the left side, where it
gives the percentages, that scale changes from slide to slide. So bar graphs can be very
misleading if you just use those alone. So you're better off just looking at the
percentages in the tables, but this gives kind of a ... a visual of what's going on there.
Hayek/ What's... what's the term on the far right of the screen? The...
Barnum/ Oh that series I? (both talking) That's, uh, when I put it in Excel and didn't... didn't
title it. That's what ... that's what it called it by default. Yeah. So, here's beat 5 now.
And, so beat 5 ... it's not quite as stark, the increase, but it ... but it, you can see from 2010
through 2012, there's definitely an upward trend, and if you looked in the earlier years,
it's a little more ambiguous looking at the graphs. And then beat 3 was the other beat
where we found, uh, a change in disproportionality, and I think, uh, this one was more of
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 10
an anomaly of 2010. Looks like there was a large spike in 2010. I don't know why. And
then, uh, the numbers started coming down in 2011 and 2012. So it looks like most of
the action's on beat 2, to me. I ... I also put up here ... the other two beats, just so you can
see, and you can see that they show sort of a mirror image to what we have been looking
at with beat 2, beat 5. There wasn't a dramatic increase in disproportionality on beat 1 or
beat 4, which also lends credence to what I'm going to talk about here in a second, that
that baseline isn't really jumping. It isn't changing, at least not citywide. It may be
changing on beat 2, because that's where we're seeing all the disproportionality, but
citywide ... if it were changing, we'd expect all the numbers to be higher on each beat in
the later years. And so you can see there that on beat 1 and, uh, so beat 1 is the top ... top
left. And beat 4 is the bottom right. You can see that actually
disproport ... disproportionality decreased, the level of disproportionality decreased. So,
do you have any questions yet before I move on to the next part here? Cause I ... I'm
throwing quite a bit at you.
Throgmorton/ (mumbled) ...just say, if you can go back to the previous slide...
Barnum/ Sure!
Throgmorton/ That chart is surprising. I didn't see ... I mean, I know the chart wasn't in the
report, but I didn't read the numbers that way.
Barnum/ And part of the reason why it ... again, this is ... part of the problem with the bar graphs.
If you'll notice, so look at the numbers, uh, on the left -hand side, the percentages here.
You can see they're, uh, they're lower than when we look at the numbers for beat 5. See
how much higher the numbers are? And so it's distorting that. It's ... it's a little bit of a
distortion. So you went from essentially 13% to 20 %, and then it came back down to
roughly 13 %. So for some reason in 2010 there was, uh, a blip in ... in, uh, beat 3, and the
numbers increased, but they came back down, and that could have been, you know, um,
for one year's worth of data, that could just simply have been an individual officer
causing that. So, uh, you know, the way to interpret data like these is to look at it as trend
data over time, look for patterns that seem to be consistent, and there's a few patterns that
are showing up in the data. Any other questions? Okay, so I just wrote some
generalizations, and these are in the report. So the percentage of minority drivers
stopped, uh, significantly increased, um, from the 2005/2007 levels when you compare it
to the 2010 /2012 levels, and uh, and that's statistically significant, if you ... if you use
logistic regression, um, and use a ... a `p' value of .001 because ... and it's so large. It...
it's significant and the other beats are not. So, uh, significantly, you know, we can say,
yeah, there really... something did ha ... happen in the data between those two timeframes.
So the earlier timeframes there were lower levels of disproportionality and those numbers
tended to creep up in the second half. Second half of the data set. And, uh, the second
bullet point just talks about what I've been talking about! Most of that seems to be
driven by activity on beat 2, which is the southeast side. And, urn ... there's ... there's
several reasons why that occurred, which we're going to get to here in just a second.
Okay, so again, just to refresh you on the baseline. These ... this is the ... the numbers that
we had for the baseline when we are watching traffic. So we went out with our
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page I I
observers. We watched traffic on all four beats, and then we did a lot of over...
oversampling on beat 2, after we saw the trend in the data. We started to see that in 2010,
and so we wanted to know if what was happening on beat 2 was being driven by a change
in the baseline, and so the way we did that was we went out and watched a lot of traffic
on beat 2. We oversampled on beat 2, and I'll get to that in a second, what we found, but
uh, just briefly we didn't really find any differences from our earlier observations on beat
2. So the first thing that we did, uh, as we were trying to figure out what's going on on
beat 2, because we saw that increase from 13, 14% to 25, 26% in disproportionality, and
so we went and started looking at, uh, the Census data first and this table just gives you
the Census tracts for the southeast side, Census tract or a subdivision of the City, which
are, uh, probably ... oh ... there's probably ... uh, roughly 100 blocks in a Census tract.
Some of the Census tracts are outside of the ... the city limits, so um ... you know, you
have to interpret that, and the numbers given here are not for minority members now.
This is for African Americans. This is a percentage of African Americans living in each
of these Census tracts. Now... um, as you can see, let me see if I can... sharp
enough ... right here is ... Highway 6 and that kind of divides north and south on the
southeast beat. So we can kind of talk about what's happening on the north side of the
beat compared to what's happening on the south side of the beat. And on the south side
of the beat is where we're going to find higher percentages of African American, uh,
residents, and drivers, when we go out and observe. So, um, on the north side, the
percentages of African Americans are lower than 10, lower than our baseline. Quite a bit
lower, and on the south side they're higher.
Mims / When... question.
Barnum/ Sure!
Mims/ When you calculated these percentages for north and south, how were those calculated?
Barnum/ Uh, you go to the ... you just look at the data from the Census tract, and it (both talking)
Mims/ Okay, so you didn't average the percentages from ... you didn't average these percentages.
You actually used the raw data ... for the entire (both talking)
Barnum/ ...I used, yeah, off their web site. You can just ... you can bring up a Census tract, click
on it, and it'll give you... it'll list out how many African Americans and then the total
population, and that's how I calculated it.
Mims/ Okay. Thanks!
Barnum/ Yep. So, since... since we had higher percentages on the south side of the beat, we
looked more closely at the south side, and now we're looking at a Census, urn ... a Census
data point known as a `block group,' and there's usually two to three block groups in
each Census tract. So a block grant is roughly 20 to 40 blocks is the way to think about
it. And these are the block groups on the southeast side. And ... so, um ... that area along
Sycamore Street and just south of Highway 6, the percentage of African Americans
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 12
residing in that block group is about 27 %. If you go just to the east of that, it's about, uh,
20 %. And to the south of that it's about 11 %, and then if you go to the west of that, it's
about 10 %. Okay. Now ... don't confuse the Census data with the baseline data. These
are just the percentage of people living there. It's not percentage of licensed drivers
living on the beat. It's not the percentage of drivers on the road on the beat. That can...
that can vary widely, uh, people travel through and so on and so forth who don't live in
the neighborhood, and so using Census data is really hazardous to try to ... try to establish
a baseline. But this gives you a feel for what the ... what the population looks like in the
different areas of beat 2, where we've seen that seems to be where the action's at. So, we
went out and we looked at traffic on beat 2, and these circles indicate the areas where we
sat and looked at the traffic, um, and so you have, uh, Scott Boulevard and Muscatine,
and then Muscatine and First, Sycamore and California, and South Gilbert and I can't
remember what the cross street of that was, is where we looked initially. And you can
see the percentages of the drivers that we found for minorities. Now we're back to
minorities, okay, so now these are people... our... our observers are sitting there looking
and they're saying is this a white or Asian driver, and if that answer is no, then they either
say well okay it's ... it's African American or it's some other racial classification, and
they ... we classify that as a minority member. You can see, uh, the numbers are lower
than what we saw on the Census data, but again, if you look at it you'll... you'll notice
that ... the percentages on the south are still higher than the percentages on the north. So
what's going on? Well, we looked a little more closely, and we looked at an area that's
bounded by, um ... uh ... Broadway and, uh, Cross... Cross Park, is that ... yeah, that's the
name of the street. Um, and we ... we ... did some observing in that area. That's not a very
big area. It's a few blocks. And about 40 to 50% of the drivers in that area were African
American. So there's an area on the southeast side with a very high concentration of
African American, and probably other minority members too, right in there. And then it
kind of disperses out and becomes ... the percentages becomes lower and lower as you
move away from sort of that, um, center, that center hub. And so, um, the question
becomes then, well ... on the north side of this beat, of the southeast beat, the ... the
baseline numbers are ... are not high. We wouldn't expect really an ... an increase in
disproportionality there because the ... the numbers are actually lower than what we were
using for a baseline. But on the south side they were a little higher, and so instantly that
tells you that well, maybe the ... the Police Department shift their focus to focus on the
south side, the south side of beat 2, and if they're doing that, then the baseline really
wouldn't apply, a 10% baseline, if you're ... if you're stopping cars where 40% of the
drivers are black, then obviously saying that you should be stopping 10% doesn't make
much sense. So the 10% baseline applies to the entire beat, okay? But if there's a
focused patrol within that beat, within a subsection of that beat, then the baseline
numbers go out the window and we can't really use `em, okay, so that's very important to
keep in mind.
Dobyns/ Question.
Barnum/ Yeah!
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 13
Dobyns/ (mumbled) ...my Iowa City geography but ... on all these sites where you're observing,
are these arterials or are they feeder streets or are they just...
Barnum/ They're mainly arterials. You don't get enough traffic otherwise.
Dobyns/ Does that corrupt the data (both talking)
Barnum/ It ... it can, and that's one of the ... that's one of the things. You get a lot of people from
outside the neighborhood ... traveling on an arterial street, yeah, and that's why you have
to be real careful with baselines. Um, but if you watched ... if you were to watch like a
residential street where five or six cars pass in an hour, you wouldn't get a large enough
data set. It would take forever (both talking)
Dobyns/ (mumbled) ...power would be.
Barnum/ Yeah, so ... um, so it's ... so that's why you're ... we're very careful with what we say
about the baseline. Okay, so ... as I mentioned to you, we wanted to know ... if that
baseline changed from 2005 through 2012, and our observations indicated that it didn't,
because we came back with roughly the same numbers that we had in 2007, when we did
it again in 2010 and 2012, when we looked at the numbers again. They were about the
same, but again the baseline, you know ... so we looked at school data, Iowa City Public
School data, and um ... the, uh, the rows that I have bolded are, uh, kind of the ones to
focus on. So, the beginning year was 2005 and what those numbers are the percentage of
African American students in each of those schools. And then the District totals at the far
right end column. And then 2007, 2008 is when our data set cut -off, and we picked it up
again in 2010, 2011. So really the way to kind of look at that is to see, you know, we saw
the jump in numbers between 2007 and 2010, so did the school enrollment numbers
during that period of time change? If you look at those numbers, they really didn't.
There wasn't much change. Um, and again, lot of school kids don't have a license, so
this is ... this is just sort of an estimate, to try to see if things are changing. But you can
see, if you just look at, uh, District total, the numbers about the same. And if you look at
all of the schools, they're relatively the same, with the exception of Wood, which
increased during the study period, but it looks like most of that increase happened
between 2005 and 2007, and not the second half.
Mims/ (several talking) ... clarification on... on, uh... on this table, there's... and I don't have the
school boundaries in front of me, but I guess I'm confused about that we've got kids that
are way down in beat 2, which is the southeast side of Iowa City that are assigned to
North Central and Northwest?
Payne/ I thought the same thing when I read this.
Barnum/ Yeah, that was a question I had before. I think in the, maybe in the very early part of
the, uh, study period, some ... that North Central has a little sliver there. So we just
included them all.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 14
Throgmorton/ Well likewise I was surprised not to see Longfellow. I ... I don't know that it
really would have a signif... significant bearing on the results that you've (both talking)
Barnum/ That might have been ... we might have overlooked that. Okay, yeah.
Payne/ But the whole point of looking at the data is to validate the idea that the population didn't
really change over the time period.
Barnum/ Right. And if you look at school -wide, 2007/2008, 16.55; 2010/2011, 16.22. So it
didn't. There's a perception in town that it did, but it didn't.
Hayek/ I thought your report indicated that there was one exception. I thought the report
indicated Wood was (both talking)
Barnum/ Wood, yes. Wood... Wood, uh, as you can see if you look at the ... the entirety from
2005 through 2010, it increased from 28.61 to 39.35, but if you look at 2007, you can see
that most of that increase, uh, was in the first part. So, it was from 2007, or excuse me,
2005 to 2007 is where a lot of the increase took place, and we didn't see the jump in the
baseline then.
Hayek/ I see what you're saying.
Barnum/ The jump came after that, in their ... in the big, according to the school enrollment, you
know, there wasn't a big jump there. That's not to say there isn't a ... a jump, a change in
the population. I can't say that for sure, but the data indicate that there really probably
wasn't ... a huge increase.
Throgmorton/ I don't know if you've had a chance to cross -check with the School District, but
that might be a... a good idea, if you haven't...
Barnum/ These reports were provided by them.
Throgmorton/ Yeah, the ... the data is yeah.
Barnum/ Yeah.
Throgmorton/ I ... I know it's not ... this isn't consistent with my memory of other data I've seen
(both talking)
Barnum/ This is actually in their, the reports that they publish. I just took it right out of their
tables. So ... yeah. Okay! So, um ... we saw that, uh, Grant Wood Elementary, there was
that increase, and you can ... I pin pointed it on the map here for you, which is ... and ... so
our best estimate about what's going on is that most of the beat on ... on the southeast
side, that beat number two, is the 10% baseline's valid. Except for the area, that little, uh,
circle area. It's a poorly drawn circle, but that little area there, it's, um ... the percentages
are higher there. The percentages of minority drivers in that area are higher than the rest
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 15
of the beat. Okay? So, anything outside of that area if a stop is made, the 10% baseline
would apply. But stops made in that area there, that 10% baseline isn't valid and it... it...
uh, it doesn't really apply. Now as we're gonna see, following 2009 when all these
incidents happened in town, the Police Department started a new beat on the southeast
side. So that's beat 2. They started a new beat, 2A, and beat 2A's boundaries are pretty
much circumscribed by what I've drawn on the map there. Okay? So the officers in beat
2A are working in an area where that 10% baseline doesn't apply. So when we see
officers who are working in beat 2A, we know that the numbers are ... the numbers that
we're going to be talking about are based on a 10% baseline, but they don't apply to
those officers, okay? They ... we can't apply ... we don't, we're not sure what the baseline
is there, but we know it's higher than 10 %.
Mims/ So where did the 20% come from?
Barnum/ That's... that's based on our observations down there, that 40% area and then what we
saw going out, and the Census data. When we looked at the Census data, that's our guess
about what's on the roads. And you know it's an educated guess, but it is a guess. I, you
know, it ... it's a fun thing to kind of do is to go out and watch the traffic, and you can do
it for yourself and just see, you know, for yourself. Just sit there and watch it and see
what percentages are on the roads. Yes, sir?
Dobyns/ What was the historical stated reason for drawing that, um, sub -area within 2? It was
drawn in that geographic area because...
Barnum/ I drew that, okay? That's my drawing based on what I ... where I think ... the population,
uh, baseline, minority baseline increased. In that area.
Dobyns/ Minority Census data. Okay.
Barnum/ Yeah. It's based on our observations, and the Census data, yes, sir. That's correct. So
this has nothing to do with the Police Department at this time. It's ... but, uh ... I drew this
independently of them, but their ... their beat 2A is pretty similar to this beat.
Dobyns/ What was historical basis for the Police Department drawing 2A at that time?
Barnum/ Well, I think the Chief 11 be able to...
Hargadine/ It was the increase in violent crime in that (both talking)
Dobyns/ Recorded crimes geographically.
Hargadine/ Correct. Uh huh.
Barnum/ So, this is just summary where we're at so far. Little more than half way done. So ... in
my view it's unlikely that that baseline changed radically during the study period. I just
don't think it did. But the level of disproportionality did change. Now when we consider
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 16
that with the fact, uh, I have the Louisville numbers there, written right there. We're
similar to what we saw in 2005 through 2007. What you would have to believe to believe
that the baseline changed is that that population remained stable from 2001 up through
2007, the end of 2007, and then in 2008 it dramatically increased for that baseline to be
that ... that divergent, and it ... in a two -year period of time it's possible, but it's unlikely
that that happened. So taken as a whole, uh, my best guess is that that baseline did not
change by much, okay? So the increase in disproportionality wasn't due to a change in
the baseline. It was due to a change in, uh, Police tactics. That's pretty much the only
other, um, option. The only other possibility for the increase. Now the ... the, what we're
gonna see is the Police did indeed change their tactics. They've been very upfront about
it. Um, and they established this beat 2A, but they did that in response to what they
perceived, and ... and what the statistics show, was an increase in crime in that area. So
they intensified patrols in that area, and this area was .... and they intensified patrols in an
area that was correlated with a higher minority population, and when you do that, you're
going to stop a higher percentage of minority drivers, and that's essentially what
happened in that area. So ... this is, uh, this is another graph that surprises people. Um,
this is the ... this is an estimate of violent crime. It ... it uses homicides, aggravated
assaults, armed robberies, and ... um, sexual assaults in the city, takes a number of those,
then it divides by the population of the city, and you multiply that times 100,000 to get
a ... a normed measure. And if you look at it over time, going back from 1999 through
2012, you can see that the crime has dramatically decreased in Iowa City during that
timeframe. It was much higher at, uh, the beginning of the century than a decade later
down in here, but you can see ... if I can keep this working... right in here is 2008, 2009.
There's a spike. There's that spike in crime right there. So it was going down, down,
down, down, then it spikes upwards, and sometimes when you have a downward trend
and then a reversal, it actually is perceived as worse in the public than just kind of a
steady state of a higher level, because you have a change. It seems different. It seems
like things are changing. And if you go back and read the news accounts of that, from
that timeframe, and by the way we did that. We ... we looked at all the news stories on
violent crime from 2002 through... 2012. And they increased dramatically. (laughs)
They ... it's almost a mirror image of what you see there. There was very little reporting
of violent crime in the first part of the decade. In the second part there was a lot, and it
peaked in 2008 and 2009, right around those, the Mother's Days incidents, that riot, so-
called riot. And so ... the ... now this is my opinion here. I think the Police Department
was experiencing pressure. I think there was, uh, pressure from the public, and ... perhaps
political pressure that something be done about the crime. And so the Police Department
went into the southeast side. They established a new beat. I think they probably sent
their SCAT officers on the southeast side, and ... they started doing some, uh,
enforcement. It's called hot -spot policing. Hot -spot policing, by the way, uh, the most
recent data is showing that it is an effective way to lower crime, and so they were using a
method that's... that's known to work. They went in there and uh, but they were doing it
in an area where the percentage of minority residents, and drivers, was higher, and that's
why the disproportionality occurred. So...
Botchway/ I had a quick question.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 17
Barnum/ Yes!
Botchway/ I know that's your opinion, but I ... I'm assuming based on the memo, Sam, you agree
with that opinion as well? Chief Hargadine?
Hargadine/ Yes, that's correct (away from mic)
Botchway/ Okay. I just wanted to make sure that was clear, cause I was taking you for what you
were saying. My other question, um, circles around, uh, the violent crime rate. So
obviously as you see it goes down, but I guess your data speaks to traffic stops, and so...
I'm confused in the sense that, um ... and I'm ... and also I'm asking a question I guess. So
let me ask the question first. If the violent crime that you speak about is characterized as
being ... uh, where's the...
Barnum/ Homicides (both talking)
Botchway/ ...homicides...
Barnum/ Aggravated assaults, yes.
Botchway/ Those aren't necessarily attributed to traffic stops.
Barnum/ Excellent question!
Botchway/ Okay, and so ... that's where I'm confused.
Barnum/ Yeah. No, that's an excellent question. Um, and there is no easy answer to that. But
often times, um, police departments when there is spikes in crimes, they'll go into
neighborhoods and use either what's called hot -spot policing or a zero tolerance policy.
And, zero tolerance doesn't have to be just traffic stops. Uh, it can be arrest anybody you
see intoxicated, whatever, you know, but given that... given the structure of that
neighborhood, traffic stops is the main modality that would be used to do zero tolerance
policing.
Botchway/ Okay. So then (both talking)
Barnum/ ...whether it lowers crime or not, that's... that's your question...
Botchway/ Right.
Barnum/ ...isn't it? I don't know.
Botchway/ So again, then back to you, Chief Hargadine. Is that what we did? Did we do, um,
hot-spot ... is it hot -spot policing, where there was a zero tolerance policy?
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 18
Hargadine/ I don't know that I would use the term zero tolerance. We were absolutely engaged
in hot -spot policing. It's ... it's, uh, similar to a tactic we're using right now in the
Pheasant Ridge neighborhood. Um, the perceptions at the time weren't just ... as ... as you
can see the crime's going down. There were, uh, neighborhood associations that were
lobbying the then Council. They were ... they were showing up, they were wanting
Council Members to come down to their neighborhood association members, uh, to
include with me. Uh, we had numerous, uh ... uh, meetings in the Grant Wood
gymnasium about the ... the spike in crime. So, um ... what we're seeing in this chart
doesn't necessarily correlate to the ... their perception of their quality of life at the time.
They ... and this is also the same time period that we started talking about the ... the curfew
and the need for the curfew. Uh, I can remember Matt saying it was a... it was a sacrifice
the entire city's going to make. Um, because of the issues at that particular, uh,
neighborhood, but it was the neighborhoods that ... that brought that forward. So ... um,
yeah, it ... it, there was definitely some pressure. We can't sit by and watch the ... uh,
those kind of crimes come in day in, day out. We're not doing our job if we don't do...
do something to that extent. As far as the traffic stops, there are usually when ... when
there's an incident like that, you'll get a description of a car that's leaving the area. So,
more often than not, it's going to end up resolving in a traffic stop. Maybe here,
maybe ... maybe, you know, they get `em stopped in North Liberty, but um ... that's the
first thing we ask is ... are they leaving and what kind of car are they, so, uh, now the ... the
officers that are assigned to that particular area, um, they get to know people quite well.
They get to know if there's a criminal element. They get to know if they're driving
without a license. They get to know whether they're wanted. Those kinds of things. So,
they'll use the traffic laws to their advantage, urn ... to make an arrest, uh, may ... may or
may not be traffic - related. Am ... am I answering your question there?
Payne/ Well and...
Botchway/ It did, but that made it seem (both talking)
Payne/ But...
Botchway/ ... highly ... I don't know. That was weird how you said that last part. I mean, as far
as using the traffic laws to make an arrest that wouldn't necessarily fall under, um ... what
you would call something they should have been stopped for if they would (both talking)
Hargadine/ Well, if they know someone is ... is driving while barred or wanted, then they don't
need probably cause. They ... they've already got the probable cause to make the stop.
Botchway/ All right.
Hargadine/ Um, but there's other times when if they know a car is ... is associated with trouble in
the past, and now they see it going down with an expired license plate, that's probable
cause to pull it over. Um ... it's a pretext ... um ... um ... you know, they're... they're pulling
over individuals in a ... in a high crime rate area that are known to have been ... part of the
problem.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 19
Payne / And doesn't it make sense that if you're patrolling more in a certain area that you're
going to stop traffic, you're going to have traffic stops more because... you... you have
more people seeing people speed ... you have more police seeing people speeding or like
you said, driving with a... an expired license plate or whatever, I mean, if you have
more... if you have a higher concentration it makes sense that you would... do more
traffic stops. And do less someplace else if you have the same number of officers.
Hargadine/ That ... that's, I think you can make that correlation.
Throgmorton/ I ... I'd like to follow up on the chart a little bit more; maybe Chris could answer
these questions, Chief. Um, two ... two things really leap out at me when I look at the...
the chart. One, you already drew attention to and that's the decline. So, um, you know,
if one extrapol... does a linear extrapolation of... of that trend, seven or eight years from
now there'll be no crime in this city, which you know ... I guess I'm joking. It's not
gonna be the case but, uh, something good is going on here, and it has to do with the
people of the city, and it has to do with the Police force, so ... I think we should observe
that, but the second thing is I think in the context much more important, your report
draws attention to the spike in 2007, 2008 or that... basically that timeframe. I don't see
much of a spike there. I see a spike in 2001 to 2003, a much more dramatic spike, and
yet no attention was drawn to it by the, uh, by... as far as I can recall, at least, by the
public or by anybody else. Instead there was considerable worry about the perceived
increase in the crime rate that, uh ... uh, ap ... was claimed to have occurred in 2000, 2008
...2007, 2008. I remember the time very clearly, and I actually organized a series of
events at the University of Iowa having to do with that. A lot of it had to do with the way
people talked about, uh ... uh, activities, and it had to do with the mental images people
had about where the crime was occurring, and so on, but I don't ... I just don't see a ... a
significant spike there. So there's something important there (both talking) need to pay
attention (both talking)
Barnum/ Well we ... yeah, that's... that's very interesting, uh, and ... and the media's tied in with
this. Uh ... uh, in terms of the ... the number of reports going up. Um ... in the, if you're
interested in this, just as an aside, this is ... this is what would be known as a moral panic
in the research and literature which is a ... people perceive something happening more
than it is. Now that's not to say that crime didn't increase, because if we look at this
table here, um...
Throgmorton/ Mr. Versypt was killed, I mean, things happened that we...
Barnum/ Right!
Throgmorton/ ...that we cannot trivialize, yeah.
Barnum/ But if you look at...at this, um ... table here. So the top table is the south side, and the
bottom is the north side. So the ... the south side's where the action's at, and this is in
2003. This is the crime rate in, uh, these neighborhoods that I computed. Whether it be
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 20
South Point, Pepperwood, Hilltop, Grant Wood, uh... so it's... seven, uh, 746, and you
can see it goes down. It kind of follows the trend we see. Then it spikes up in 2008, in
2009, and now it's ... it's going down again. So in the neighborhood it actually followed
what was happening in the community. If you look down below, you can see ... you can
see in 2003 there was more crime. In the ... in the north side too. And it ... it's following
down. Now, there's a couple of ways to interpret this chart. So, the Police began their
intensive patrols, and you can see after they did that, 2009, that time period, 2009, 2010,
the crime did go down after that. So, uh, you know, that ... that is one interpretation of the
data that ... that it was an effective methodology by the Police Department. Other
questions?
Payne/ How was the rate calculated on there?
Barnum/ Yeah, so ... you take, um, so let's look at 2003. You would add up (mumbled) gives
you 65. You divide that by the population, which I... in the report I give what population
I used to divide that by, and you multiply it by 100,000. That's a norm measure. Just...
it gives you a rate per 100,000 people.
Payne/ So that's the rate per 100,000 people.
Barnum/ Yes. Of all those crimes.
Payne/ Okay. So, and the ... the chart, the last chart, the ... graph that you had on the other chart
was for the whole city, and this is just for that neighborhood. That's why the numbers
are so vastly different.
Barnum/ Yes.
Payne/ Okay. Got it!
Barnum/ Yes, that's... that's absolutely correct. So here's a summary for this section, and we're
now going to move into the last section. Um ... baseline I ... was in the neighborhood of
10 %, give or take. Disproportionality was lower 2005 through 2007 than it was in the
later years. It's unlikely that that change in disproportionality which was mainly driven
by activities on beat 2 and among officers who worked beat 5, which are the officers who
are not assigned. It's unlikely that that level of disproportionality was driven by
significant increases in the baseline, in my opinion. And so that means that it was
probably due to changes in police tactics, and the police have, you know, have been very
upfront with... they... they did indeed go into the, that area of town, and engaged in hot -
spot policing, which would account for a rise in disproportionality.
Hayek/ Can you drill down on the beat 5 data, um ... I mean they're not ... they're not
geographically assigned, I mean, beat 5's this ... this floating creation essentially of... that
you did for data collection purposes, but you know, wh ... where were they spending their
time, where were the stops made, uh, and how does it get reported out if they are
normally assigned somewhere else or...
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 21
Barnum/ That's an (both talking)
Hayek/ ...not following.
Barnum/ That's an excellent question. The data set we received doesn't give the location of the
stop.
Hayek/ Okay.
Barnum/ So I can't determine where they, you know, I don't know where they were making their
stops. It's all ... we're guessing. Their ... their level of disproportionality increased...
given that ... that is a crime, a special crimes unit that's ... goes into high crime areas, I
would guess that they were probably on ... in that area. I don't know that for sure. Future
studies, when I get to the recommendations, should track where the stops are made.
Should also track where the officers are spending their time on the beats. Um, and that
would, Mayor, that would answer your question then. From this ... from these data that
can't be answered.
Botchway/ Well in that case wouldn't then the... percentage increase drastically, if you attribute
the beat 5 officers to the particular beat 2 or 2A?
Barnum/ Yes it would. Yes, if they were spending all their time say in that 2A area, then...
because the baseline is so much higher, yes, we would expect that the numbers of
minority drivers stopped would increase. Now the question is, the broader question is,
we know the entire beat is 10 %. And so the question which, uh, occurs to people is why
are the officers spending all their time in the areas where it's not 10 %. And that, I think
the Chief is ... addressed that issue in his memo. That's why the officers were not on the
rest of the beat. They were in the one area.
Payne/ So, I have a question about, related to what you were talking about, about not knowing
the location of the stop. I understand it ... if it's not provided to you in the data, that you
can't use it to track.
Barnum/ Right.
Payne/ But I don't understand why it can't be ... why it's not provided when, I mean, it says when
stopping a vehicle, officers contact the dispa ... dispatch center, who logs the stop.
They're not logging where the stop is?
Barnum/ No, they are logging where the stop is, but that would ... so if you ... if you imagine the
raw data set, that would be a series of addresses which wouldn't have any ... it wouldn't
be sorted in any way. So it's very, very difficult. I didn't receive that data, but had I
received it, it would have been extremely difficult to sort that by beat and so on and so
forth. There's some ways to do this, moving forward, which I'd be happy to share (both
talking)
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 22
Payne/ GPS plopped a dot on a map?
Barnum/ Yeah, that type of thing (laughter) Um, I also do this work for Davenport, and they...
they have an interesting way of doing it. Uh, which is very ... very effective and um, so
I ... I'd be happy to share that with you when we get to the recommendation portion.
Okay, so now...
Payne/ I guess my point was it seems like we're making a lot of guesses when we theoretically
have the data, that we don't have to necessarily guess.
Barnum/ I'm not sure that those data are in a form though that can be analyzed the way they're
currently being collected. So, yes. That's, you're right. Uh, it doesn't have to be a
guess. It can be a certain ... so, if you knew where the stop is made, you don't care what
beat the officer's working so much as where the officer's making the stops, and which
does change things.
Payne/ Yep. Thank you.
Barnum/ Yes. So now what we're looking at is what ... what beat the officer's assigned to and
just assuming they're on their beat. Okay, I hate to do this to you. I have to do a little
math here. Um, so the ... the, urn ... the next part we're going to talk about individual
officer data, and we use something known as a disparity index, which is similar to an
odds ratio. An odds ratio is just a way of comparing two different groups to each other...
on the same scale. So, we know that the baseline, we're just going to assume that the
baseline is 10 %, right. 10% minority, 90% non - minority. And so what you do is ... for a
given officer is you take the percentage of stops the officer made that were minority
members and divide by that 10% baseline. That gives you a number. So let's say an
officer made 100 stops and 10 of those stops were minority members. So, 10 ... that's
10% of the stops for that officer were minority members. We would divide that by the
minority baseline. Are you with me so far? So ... 10, 10% divided by 10% is 1. Then,
oops, excuse me (noise on mic) And we do the same thing for the number of non -
minority stops. So for that officer it would have been 90 of this officer's stops would
have been non mi ... non-minority members, and the baseline is 90 %. 90 divided by 100's
90 %, so 90% over 90% is 1, so we have two ratios, that's why it's an odds ratio. 1 over
1. So the odds are the same that a white and a black driver, just to simplify it, have equal
odds of being stopped. So a ... a value of one means that the odds are the same that one
group isn't more likely to be stopped than the other, but if you say ... let's say with this
same example this particular officer, who still made 100 stops, but this time 20 of the
stops were, uh, minority drivers. So now 20% of the officer's 100 stops are minority
drivers, but the baseline's still 10 %. So the numerator, the top number, is a 2 now, right,
and the bottom number is 80 divided by 90. Whatever that is, and then if you divide
those two things, that'll give you the odds ratio. And so what you have is numbers larger
than one indicate disproportionality. And as the numbers get larger, that indicates more
disproportionality. Okay? So you want to look for numbers bigger than 1 and because
our baseline isn't perfect, I would give it some flexibility here. I wouldn't just strictly
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 23
interpret a number greater than 1 as being meaningful. I would start looking at numbers
say maybe bigger than 3. So that ... that's indicating that according to a 10% baseline, the
odds are three times as great that a minority driver would be stopped in comparison to a
white driver. Another way to think about that is that the odds increased by 300 %. So
it's, you know, and as you move up, um, you know, so as those numbers get larger, that
indicates more disproportionality.
Dobyns/ So by suggesting that you have to have a 300% increase, (mumbled) look at the
numerator, um, you pick a fairly large number cause you realize, um, all the assumptions
placed in this, and the bias of hot - spotting.
Barnum/ Yeah, that's why ... that's why when we look at beat 2 officers, we cannot ... we cannot
apply the disproportionality index to them. Because ... the baseline isn't right for them,
isn't correct, and ... and you hit on something that's very important. The more stops an
officer makes, the more confidence you can have in this index. So if you have an index
based on 20 stops, I wouldn't put much credence in that. If I had one based on 200, 300
stops, then you're getting more, you know, to an area where you can have more
confidence in it. So, that's... that's, so when you're looking at the data that I'm about to
present, don't just look at the index. Look at the number of stops, as well. Now these
data are de ... de- identified, um, so the officers aren't shown, but you ... you'll be able to
see on there what the distribution of all the officers look like for the entire Department.
So this is 2005. The bottom left box are the disparity index values for the officers with
the highest values, and you can see what beat they worked on, and the number of stops
they made. So that second officer on there was on beat 2 and made 263 stops, and the
next one made 508 stops. So those are fairly high number of stops, um ... and so
confidence increases more so than say that top number of 51 stops. That's the way you'd
look at it, but if you look up at top at the graph, this kind of...well I think I ... there we
are! So, each of these little dots is an officer, and as you move along the horizontal axis,
these are the odds ratio values. So, if you look at number one here and if you counted
those officers, that's how many officers had a odds ratio of about 1. (mumbled) how
many would that be? Two, four, eight, yeah! Yeah, so eight officers in 2005 were right
dead on. Some were below! Below 1, which means that whites were stopped at higher
percentages than we'd expect ... in the population. And so the way to look at this is look
how all the officers are grouped together, given that the baseline is not perfect. You look
for outliers, and you look for trends over time. Two things you want to look at in the
data. Outliers and trends over time, and these ... in this data you can see that there's kind
of two, two of these officers that are a little bit higher than the rest. Uh, one has an odds
ratio of almost 5, and that's that person with, uh, 51 stops. So the stops aren't very high.
The ... the number of stops aren't very high. And so, you know, we don't know. That
...that literally with that few of stops, it just takes a couple when you don't have many
stops to really spike the number. So...
Mims/ Can I ask for clarification on terminology?
Barnum/ Uh huh.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 24
Mims/ I got confused in here. You start talking about disparity index when you're talking about
stops.
Barnum/ Yes.
Mims / And then you went on and talked about standard odds ratio to evaluate pro...
disproportionality dealing with the outcome of a stop.
Barnum/ Right.
Mims/ Then we got to the tables that are talking about stops, but you're labeling the bottom of
them as odds ratio.
Barnum/ Uh, yeah (both talking) I'm sorry if that ... that's a ... that's an artifact of SPSS, the ... the
statistical package that I use. It stuck that on there. That's a picture.
Mims/ Okay.
Barnum/ I couldn't take it off. That's just the name of the value I used for the variable. I'm
sorry about that. That's ... I ... I, yeah, I should have pointed that out. So no, this is the
disparity index. These values.
Mims/ Okay!
Barnum/ That's a very ... thank you for pointing that out.
Mims/ Thank you.
Barnum/ And sorry for the confusion there. Yeah, so this ... this is the disparity index here.
Mims/ It's a statistical package (laughs) (several talking)
Barnum/ Yeah.
Dobyns/ Other than a small amount of data points and (mumbled) by the, um ... people who are,
um, you know, at 4 and 5 odds ratio, any other distinctive characteristics of those
officers? (mumbled) ...assigned to a hot spot or ... I don't know.
Barnum/ At this time, beat 2a hasn't come into existence yet. So these officers are working beat
2, uh, the ones that are on ... you can see most ... most of the officers are on beat 2 that
have the high (both talking) have the high values. So the way I would interpret that is the
officers are probably making more stops south of Highway 6 than north of Highway 6. I
don't know that for certain though. So the mean average, when you... calculate all these,
and this ... this mean isn't weighted, so it's not that good of a mean, but it gives you a
rough and ready estimate. Is ... is ... it's lower than 2. So in 2005, there wasn't much
disparity, is the way I would interpret that. That's lower in 2 and uh... you can see
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 25
they're all clustered around 1 for the most part. Here's 2006. It's creeped up a little bit
in 2006, the clustering, um, is ... there's still a lot around 1. Kind of stretched out a little
bit. The highest officer has an odds ratio of 6, but you'll notice only 25 stops. And so
that's ... that really isn't enough to, uh, put much credence. The next one though has 776
stops. That's a lot of stops. And so that number, that number is probably more reliable,
more valid. You can see that the mean average then weighted mean is 2. It went up from
1.71 to 2. Still at this point, uh, remember I said that I look for values larger than 3, so
this still doesn't look terribly, um...
Dobyns/ I'm trying to think about that officer with, uh, 776 stops. Um, under current, uh, police
operations. Um ... and this is in 2006, so this is before that bump. Uh, we had no
perception, I think, at that time historically of a hot spot. So, perhaps in 2008, 2009, but
in 2006, I don't recall any reason why, um ... the, uh ... police, uh, leaders would have told
that officer to perhaps drive south of Highway 6. Um, that officer may have chosen to
(noises on mic, unable to hear speaker) Highway 6, or he was told to drive south of
Highway 6. Since there was no reason that he was told to do that in 2006 that I can
recall, that officer chose to drive that route.
Barnum/ Uh...
Dobyns/ (both talking) ...he was assigned there, but ... but normally on police, um, are they
assigned ... if...if you're assigned 2, I mean ... where do you go? Where are your haunts?
I mean (both talking)
Barnum/ Right.
Dobyns/ ... are you told that you need to randomly drive around all of 2 and be intentional about
that?
Barnum/ I can only speak from my own experience. I was a police officer in Cedar Rapids for
25 years. Um, your ... we were left with a great deal of discretion on where to go ... on our
beats. Now, let me say this ... the, um ... you need to follow this officer over time, because
it may be the next year this officer has a low odds ratio. So something could happen ... I
mean, there... you... you know, maybe this guy or I shouldn't be gender specific, but it
was probably a male. Setup to run radar in a certain spot that just happened to have, you
know, I mean it ... it's possible, you know, so you have to look at the numbers as trend
data more than snapshots.
Dobyns/ And your officers are identified from year to year so I can't follow (both talking)
Barnum/ Right, right!
Throgmorton/ Could I in ... introduce a little bit of different history. Um, you know, I was on this
Council 20 years ago and at...at that time, I can re ... recall very, very clearly various
people, rightly or wrongly, complaining about the Broadway neighborhood. And about
how there was, you know, there... troubling people down there and an increase in crime
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 26
and that kind of stuff. So the ... the tale goes back at least 20 years. And I don't know if
that had any effect on the, you know, the data you're pointing to, Rick. I have no idea
whatsoever but...
Barnum/ (both talking) That's an interesting point because you've jogged my memory now.
When I was reading about the SCAT unit, it was actually... according to what I read, I
don't know if this is accurate. I read this in the newspaper. It was actually developed in
1999 as a response to Br ... Broadway problems. So that's consistent with what you're
saying. So ... it's possible. Here's 2007. Uh, the average is dipped again, back down
below 2. Most of...of the officers are again below 2. Um ... with the exception of the one
individual.... you'll notice there's an officer on beat 1 that has a fairly high odds ratio.
(clears throat) And so ... for the most part, these data show that the levels of
disproportionality are pretty low. You ... you have some officers that you'd look more
closely at. And that information was provided. So there... there's more detailed, uh,
information that's ... those officers have been looked at more closely but um, again, what I
suggest is that we look for the trends. And so we jump ahead to 2010. Now ... the pretty
colors up there are the officers who were assigned to beat 2A. The red officers were
assigned to beat 2A full time, and then the officers in green, which was an unfortunate
choice by me (laughter) It's very difficult to see! Uh, I'll read it to you. Uh, they
worked that beat part of the time, and I don't know how ... how often they worked it,
again, but they were assigned some of the time to that beat, and that would ... that would
dramatically escalate their disparity index values.
Dickens/ Well I know I ... I rode with one of the officers in the fall of 2009, or two different
officers, and they were... one was assigned to downtown. The other was kind of a
floating, but if they were called, they had to go down to the other beats. So...
Barnum/ Yes.
Dickens/ I mean, they were pulled down there for... for different reasons, and it was all call
based, so it wasn't always they're just going down there looking. It was a lot of it was
based on calls (both talking)
Barnum/ Yes, and that's quite possible too, yes. Absolutely! Um, and so the values that are hard
to see are, the first one is 6.14 on beat 2 with 69 stops. The one below that is 6.03 on beat
2 with 137 stops. And then, uh, if we drop down to the one below 5.31. That's 4.91 on
beat 2 with 266 stops. And finally, well, we have two left. 4.42 on beat 2 with 367 stops,
and 3.78 on beat 3 with 495 stops.
Karr/ Chris, this is the same as page 38, in the book.
Barnum/ Uh, I don't have my (both talking)
Karr/ Yeah, it's page 38 in your book.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 27
Barnum/ I don't think we have anymore green the rest of the way! Nope, we don't! Okay, so
here's, um ... here's 2011 now. Um, that top officer with an odds ratio of 9 only has 22
stops. That's not nearly enough, uh ... literally 3 or 4 stops can dramatically change that,
uh, the next two though have a fair number of stops. Uh, 7.43 and 6.88, but both those
officers are on beat 2A! So we would expect to have high numbers for those officers.
That's not surprising and it doesn't really mean anything. You know, because they're...
they're told to be down there! That's where there's higher percentages on the road, and
so kind of the first three I wouldn't pay much attention to, and that goes for the graph up
above too. But if you look at that graph carefully, it's interesting ... uh, because you see
that the mean average has bumped up to over 2 now. It's 2.3 1, and the clustering doesn't
seem to be around 1 anymore. It seems to be closer to 2. And so it seems like there's
been a shift. Do you see it? Uh, let me go back here and show you this one. So... you
can see the clustering's kind of around 1 on this one. And now ... we're here, and 1's not
listed on there. Again, that's SPSS, urn ... but the clustering seems to be around 2, and
that's reflected in the mean. So it looks like the disproportionality has increased, and
again we know if you just look at that you can see where it's coming from. It's coming
from beat 2 mainly. In 2012, uh, so that top officer has 55 stops, and then we have a beat
2A officer, and then we have, uh, a beat 5 officer who has 52 stops and then you can see
so it's 4.37, and the mean's about 2.32 which is, uh ... little higher than what it has been,
and you can follow that down through and look at those, and the clustering is still fairly
close to 2, and ... and so it's ... it's just, what this is showing is the same thing that we saw
when we were looking at the department -wide data. When you look at the individual
officer data, it's ... it's similar to what we saw before, but what you're seeing here is that
maybe that disproportionality, that jump from ... from, um, 14% to 18 %, a lot of that
might have been driven by beat 2A activity. It's quite possible. But again, I can't tell
you that for certain cause I can't tell you where the stops were made. So this is just what
I said, essentially, uh, this next page. I just said this, uh, the level of disproportionality
went up, mainly driven on beat 2, perhaps beat 2A. We know it was higher in 11 and 12,
10 through 12, than it was in the previous years, in the first part of the study. So just so
we had ... we have this kind of study state from 2001 through 2007. Then this spike in
crime occurred. There was a downward trend in crime, but it did reverse, kind of like the
J curve where there's a lot of research on when you have a J type curve perceptions of it
...of it being worse than it really is occur. Uh, the Police changed their tactics in
response to that and that drove some disproportionality. I think...
Throgmorton/ Interesting thing about data though, you know, when you have variation about the
trend line...
Barnum/ Yeah.
Throgmorton/ ...you would expect to see some variation; you'd expect to see some declines that
drop below the trend line. You expect to see some increases above the trend line. The
question is, what's the trend line?
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 28
Barnum/ Yeah, that's the question! Yeah, that's the question .... I, you know, and ... and for
disproportionality, that's still an open question. What the trend line ... what the trend line
for that is. That's... that's, you know, that's still an open question.
Dobyns/ Chris, I'm ... trying, I'm comparing 2 and 4. Um...
Barnum/ I'm sorry, I...
Dobyns/ I'm comparing, uh, 2 ... zones 2 and 4.
Barnum/ Yes!
Dobyns/ And I'm thinking of, um, what happened with the (mumbled) the perception of
increased crime in, uh, area 2 (mumbled) 2007, 2008. Um ... the, uh, disproportionate
index went up in 2.
Barnum/ Following that.
Dobyns/ Following that. There was a hot spot that was identified in 2, which I think affected
perhaps police operations. There were no hot spots in zone 4. Was there a similar
congruent increase in, uh, the disproportionate ratio in zone 4 that same time period?
Barnum/ No. Not in that timeframe, but the question begs is now in zone 4. Now in zone 4,
uh...
Dobyns/ Okay.
Barnum/ Pheasant Ridge is over there. And it'll be interesting to see ... what happens, because
the perception is there's a big increase in crime in zone 4 right now. Would you agree
with that?
Dobyns/ Yeah, I mean, I'm just thinking of your recommendation before... (both talking)
Barnum/ I think that the perception in 2009, you know, so if you use this level of crime kind of
the feeling about crime on 4, the feeling on beat 2, and the Chief can answer... probably
speak to this with ... it was much ... well, I shouldn't say much, but it was worse at that
time on beat 2.
Dobyns/ No, I ... I'm just trying to figure this through ... when I take a look at that data I'm
thinking that if your number goes up, if it's a disproportion ... if you're told by your
captain to go to this hot spot, and there happens to be any number of differences, but
we're measuring the percentage of, uh, African Americans in that district. If you're told
to go there, uh, you're going to get more stops and it's going to be disproportionate on
the basis of geography. It just so happens that there are more African Americans in that
section, as well.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 29
Barnum/ That's correct.
Dobyns/ Can you control for that through regression analysis?
Barnum/ If I knew where the stops were being made at I could.
Dobyns/ Right. Yep.
Barnum/ Yeah, we could...
Dobyns/ And we don't.
Barnum/ We don't.
Dobyns/ Yes. Yep.
Throgmorton/ But one can also look, tell me if I'm wrong about this, one can also look at
variation in the dis...
Barnum/ Disparity index?
Throgmorton/ What's the term? Disparity index. From officer to officer, in the same beat.
Barnum/ Yes.
Throgmorton/ Or at least even in the same sub -beat, like 2A.
Barnum/ Yes.
Throgmorton/ So one thing I see, and tell me if I'm wrong about this, is considerable variation
among officers.
Barnum/ Yes, absolutely.
Throgmorton/ All right. So, and like ... where's this one point on, uh ... um, on ... on page 49 the
report says several officers not assigned to beat 2A ... not assigned to beat 2A or SCAT
demonstrated high levels of disproportionality in comparison to their colleagues.
Barnum/ Right.
Throgmorton/ So, for me that's a huge ... that's a hugely important signal. And it's one I'm sure
the Chief is taking, uh, totally to heart, that ... that there seems to be some significant
variation from officer to officer, and we don't really want that, I think. Tell me if I'm
wrong.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 30
Barnum/ Well it ... I, I mean, urn ... you might have an officer, for example, who is disengaged but
wants to keep his boss off his back. And so he goes out and he's ... he stops, let's say he's
stopping cars for one headlight, way out on north Scott. Out, you know, just trying to get
a couple tickets a night but not wanting anything else to do. That officer was going to
have a very low level of disproportionality. He's going to be stopping mainly white
drivers. And, so, I mean, there's lots of reasons why the variation can occur, not just the
one way that indicates bias. I mean, the variation can be below and above. Now I did do
some statistical analysis, it's in the appendix if you're interested. It's hierarchical linear
modeling.
Throgmorton/ And I couldn't track that (laughs)
Barnum/ Yeah, yeah, it's, urn ... and that does look at officers from year to year. And what it
showed was the values did increase over the time period when you control for beat. Just
to put it in ... um, plain terms. So ... what you said is true. So, this gets at the questions,
the question you just asked. And, um, this is looking at now only beat 2 officers, and this
is the variation for the officers working on beat 2. And so you can see that the top
officer's value was 7.41 and ... but we had another officer that was at 1.29, and then we
had lots of scores in between; pretty wide variety. And so it...it does beg the question is
why ... why is ... is the one officer with the lower value. Now in these data it's apparent
why. Because the red are officers assigned to beat 2A. And so we would expect them to
have higher values, but we still have a pretty good range when you look at, uh, 4.22 down
to 1.29, but then we look at the number of stops for that ... for the officer that had the
value of 4.22 and we can see that the number of stops aren't great. And so a lot of times
variations in data, there's really good reasons for it ... that don't indicate bias. And in this
particular instance when you look at it, urn ... it's ... it's quite interesting. That we have
one officer on beat 2A that has a relatively low number. And that's interesting. Actually
that's, to me, that's ... from a statistical standpoint that's a little more interesting. Uh, I
don't know why that occurred. Again, that could be a statistical anomaly but ... that's
something that should be watched over time, as well. And so here we are, um, again now
SPSS tells us this is beat 2, but it's beat 5. It's beat 5, and uh, you can see that from the
beat chart. It tells you that (mumbled) pay no attention to the man behind the curtain type
of thing, but it's beat 5, and you can see that the values there, um, so some of the officers
had high values and some low. Again, and ... and I'm going to go through these rather
quickly here cause this gets kind of repetitive.
Hayek/ Chris, can I just ask you a quick question?
Barnum/ Sure!
Hayek/ When you show the disparity index values and you ... under like a beat 2 chart, within it
you break down between 2 and 2A, but are the baseline values similarly... distinct within
that chart, or are they all using the beat 2 chart?
Barnum/ They're all using that 10% baseline.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 31
Hayek/ Or the, yeah ... thanks!
Barnum/ Yeah, yeah, that's a good question. They're using that 10% baseline, but ... and that's
what's driving those numbers so much higher for the beat 2A officers, and that's why we
can't really evaluate them using that... using the numbers, but if I switch numbers then
we'd be comparing apples to oranges, because we don't know for sure what the baseline
is in beat 2A. Now we could find that out but it would take some more work. So here's
11, the same thing for beat 2 officers. Again, we have some dis ... uh, some variation.
And here's beat 5. And I'll do this one more time for 2012. Here's beat 2. Now again
now if you look at this, this is ... this is interesting because if you look at those bottom two
down there, they're right at 1. On beat 2. And they're making quite a few stops, and
then, again, though we have some high values. So we have officers that are different and
what would be interesting to do is to, and actually I did do this, um, you know, to look at
the officer's characteristics, years of service, and so on and so forth. Plug it in there.
Payne/ So this chart for year 12 ... the second line, the 5.59 says 2A on it, on the printed...
Barnum/ Yeah.
Payne/ It's the only one that says 2A.
Barnum/ Yeah.
Payne/ That means there's only one officer assigned to 2A?
Barnum/ Yeah. The ... the number of officers assigned to 2A decreased after 2010. (several
talking) I don't know why my color didn't transfer when I copied and pasted, but it
didn't. So in 2010, right after that, right after the spike, that's when a lot of officers were
on 2A, and then they have kind of pulled back since then.
Mims/ When you've done this kind of analysis, um ... is it possible with the data to look at the
time of day? Do you typically see some differences in the data based on the, um, when
the officers are on duty?
Barnum/ Um, yes, but there's um ... it's a little more complicated than just day and night, because
often times more senior officers work during the day, and when you get an officer, like I
was, who ... I never worked days, by the way, but ... towards the end of my tenure, I
wasn't as ambitious as I was at the beginning of my tenure. (laughter) And ... the, uh, so
there's a confound built in there that could be explaining the differences... you see a
difference between day and night, but it might really be due to seniority.
Mims/ Okay.
Barnum/ Do you see what I'm saying?
Mims/ Yep.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 32
Barnum/ So there's ... it's a little bit more complicated than just that.
Mims/ So looking at just... officers on the same beat, there's a lot more variables in there that
potentially could be analyzed.
Barnum/ Right.
Mims/ In terms of the beat, the time of day that they're on duty, as well as their seniority, etc.
Barnum/ Their race and their gender, yes. All those things.
Payne/ Which would have made this report 500 pages instead of 83.
Barnum/ And it was already ... it was already eye - glazing, as it was. (laughter) Okay, so ... now
we're finally to stop outcomes. So all, so far what we've been talking about is stops.
Comparing percentages of stops to the baseline values. Now we're on more solid footing
though because we're ... with stop outcomes, because now you don't have to worry about
the baseline. Cause now you're ... you compare the percentage stopped white versus
percentage stopped black just to make it simple. So how many, 50% of officers stops
were white drivers, 50% were black, but the officers writing, urn ... almost all of the black
drivers tickets but none of the white drivers. So you can look at it that way. It's a ... a
little cleaner that way. You can com ... you can compute an actual odds ratio. And...
Mims/ Let me clarify real quickly though, cause in the ... you're saying black, but in here it's
looking like white and Asian and minority.
Barnum/ It's still minority. I just use this in this example as a...
Mims/ Okay! Okay!
Barnum/ Just to make it...
Mims/ Thank you.
Barnum/ ...quick. It didn't make it any quicker though. (laughter) So...
Mims/ We did read the report! (laughs)
Barnum/ Yeah! Yeah, I can see! So, um ... again now, values of an odds ratio greater than 1
indicate disproportionality. Values of an odds ratio below 1 also indicate
disproportionality, but it's disproportionality for non - minority drivers. So what this tells
you for... in 2005, the Iowa City Police, based on odds, were more likely to write a ticket
to a non - minority driver than to a minority driver. And by a ... the odds increased by
about 38 %. That 1.38, so about ... the odds were about 38% higher in 2005 that a ... a
white driver as opposed to a black driver would receive a ticket. Okay? Uh, now arrests,
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 33
we see that number has flipped. So the odds were about 2.54 times, 254% greater, that a
minority driver would be arrested as the result of a traffic stop, when compared to a non -
minority driver. And that, you know, so you're saying, well, then the ... the odds are
about two and a half times higher. Now, there's, again, it's not quite as clear as it might
first seem, because there's lots of variables. They can ... they can come into this that were
not controlled for, uh, in this analysis, because the data weren't provided, and that's
something that should be done down the road. We don't know the reason why the person
was arrested. That's something you'd want to know. There may be, I'm not saying there
is, but there may be differences between categories in terms of, uh ... uh, different types of
offenses. Uh, warrants, driving while suspension, and things. Things where the officer
has very little discretion, and if there are, it will show up in the odds ratio. Now that's
something that can be found out if we know why the person's being arrested. So if you
have high discretionary, in other words the officer has a lot of choice. If you have types
of high discretionary types of arrests where you find disproportionality, that's much more
telling than low discretionary arrests, like a warrant. A warrant's essentially an order
from a judge, telling the officer that an arrest must be made. The officer doesn't have
any choice. And if that is correlated in some way with, uh, demographics, then that's
going to show up in your data. So that's just one thing to keep in mind about arrest data.
Now these are search requests and I'll talk a little bit more about this. Search requests are
data that are for what are called non -PC, or voluntary searches. These are searches where
the officer does not have a reason to search the car, and so the officer comes up and says,
"May I search your car ?" And the driver has the option of saying no. And ... if the driver
says no, the officer cannot search the car. Can't do any ... can't sanction the driver in any
way. There can be no element of coercion. Okay, that's... that's the idea be ... behind a
search request. Now, to track these data properly, the data must be tracked when the
officer asks the question. Not when the driver gives the response. And what we found is,
and we found the same thing in Davenport. What officers tend to do is they check that
box after they've asked the question if the driver says yes, but never check the box if the
driver says no. Which is a big problem, and ... and I mean, you know, and it's just... you
know, because of the murkiness of...of what it is and so on and so forth, but Davenport
spent a lot of time fixing that, and it takes... it takes some time to do it. But to track that
properly, you need to have that ... you need to have the officer tick the box whenever the
question is asked, and the officer should only tick the box when there's no other reason to
search the car. What often also happens is the officer will tick the box, even when they
have a reason to search the car. So the officer walks up to the car, smells marijuana. "I
smell marijuana in your car. Do you car if I look in it ?" And the driver says, "No, I
don't care." So the officer will come back and tick that as a voluntary search, but that's
not a voluntary search, because the officer, by smelling the marijuana, really can search
that car either way, and the officer's probably going to search that car, whether the off...
whether the driver says yes or no. Okay? So ... so in order for these data, in the future,
there should be some training for the officers so they know how to check that box
properly. So what I'm telling you is, when we look at these data, cause what they're
going to show is that minority drivers are asked more than non - minority drivers by
officers, if their car can be searched. But the hit rates, or what officers find, are actually
higher for white and Asian drivers, non - minorities, than they are for minority drivers. So
what the data are going to show is that the officers are asking minority drivers, "Can I
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 34
search your car," more than they're asking non - minority drivers, but when they do
conduct the search, they find something more for the white and Asian drivers than they
do for the minority drivers. Which is, you know, problematic because you're... you're
looking... you're asking people to search when you're not finding as much. You know,
you would think you would go where the ... where you're going to find something more,
right? But again now, the data for the request may not be good. That's something that
needs to be tracked more closely as we move forward. If you decide to move forward
with it. Did I make that clear? Are you following me what I was saying?
Payne/ Garbage in is garbage out.
Barnum/ Kind of, yes (laughter) Yes.
Mims/ But isn't this similar to national data too? It seems like I've read this somewhere else that
it's...
Barnum/ Yes, and (both talking)
Mims/ ...more for minorities for searches and yet when they're getting the hits it's a higher
percentage on non - minorities.
Barnum/ Right.
Mims/ Nationally as well.
Barnum/ That ... that's correct. Now there is a little bit more to that too, um, so just the ... the hit
numbers are higher, let's use just whites and blacks cause that's where most of the data
are at. So the hits are higher for whites, and if we're just looking at drug offenses, let's
say. The hits are usually always higher for whites, but they find that the quantities are
higher for blacks. So although the hits numbers are higher for whites, so ... so it's, you
know...
Mims/ Yeah.
Barnum/ ...so that can enter into it, you know. I mean, you ... but that ... that's one of the comp...
you know it's a complex issue to look at. It's not... straightforward.
Hayek/ (several talking) So two minutes ago you ... I think you indicated that, um, that ... in
probable cause situations the search request box shouldn't be checked off. Do you say
that like as a... as a data collection comment or as a... as a tactic, because I could see
situations in which ... you know, I mean, I'm an attorney and ... and there are evidentiary
issues that you're maybe trying to hedge against.
Barnum/ Right, and I think that's exactly (both talking)
Hayek/ Is that your comment or is it more of a statistics issue?
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 35
Barnum/ This is exactly what is happening. I think you hit it right on the head. I think the
officers are looking at this from court, from a standpoint of going to court. Well, the
county attorney likes it; I have probable cause on this, smelling the marijuana, but they
like as much as they can get, so I'm also going to ask the person, `Can I search your car?'
And I'm going to tick the box that way. Now, what I suggest moving forward is, they
can continue to do that, but they shouldn't tick the box. They can note in their report, the
report that goes to ... to the judge, and you know, to attorneys. They can note in the report
I also asked for permission, but you don't need to tick the MDC box because that's not
going to show up in court. And so for recordkeeping...
Hayek/ Okay.
Barnum/ See what I'm saying? So they don't really need to tick that box in the police car. They
can still ... uh, write in the report that they smelled marijuana and asked the driver if I can
search your car, but then they would ... they wouldn't say yes on the MDC box that it was
a voluntary search, cause it's not a voluntary search.
Throgmorton/ I'd like to ... I'd like to follow up on that, if I could, uh, to see if I'm understanding
because I'm not entirely sure that I am, but I want to offer an interpretation and see, uh,
what you think. So ... on page 8 when you're ... in the introductory part when you're
discussing this topic, uh, you write, um, the average odds for consent searches were about
three and a half times greater ... uh, that ICPD officers would request a search from
minority drivers compared to others. This despite hit rates that were actually lower on
average for minority drivers. Wh ... when I read that, the way I interpret it is, our officers
are more likely to request searches from minority drivers, even though they're far less
likely to find anything... uh, criminal... criminally related within the car. (laughs) And
that's a pretty bothersome, uh, way of thinking about it. I ... am I right in that
interpretation or not?
Barnum/ This ... this particular measure ... this estimator right here, is really excellent for finding
bias. This is ... this is the one ... when you ... when you drill down and you're looking for
bias, this is where you look. Uh, unfortunately for these data though, the officers are
ticking the box incorrectly. So I wouldn't ... you can't, I mean I ... I don't feel comfortable
saying that they really are ... that there is these ... these differences. That's something that
can be looked at, but ... if you just think about from like a rational consumer or a fisher...
a fisherman. You know, you drop your line where you think you're going to catch the
fish. So you know, if you're constantly dropping your line in the fishing hole where
there's no fish, it does ... it's not rational, right, you know, so ... the black pond is over
here and the white pond is over here, and the few times you throw your rod over here,
you're getting... you're pulling fish out but you continue to go over here. That ... that
would be a ... a pretty clear indicator of bias. But for these data now, because we don't
know for sure, uh, we did a survey of the officers and it came back about 50% of them
were mismarking the boxes. And we don't know where they're at in the data, and so, uh,
moving forward though, this is a good way to track it. One other good thing to track,
when looking for bias, is when an officer asks somebody to step out of the car. So ... the
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 36
officer walks up to the car and says, "Will you please step out of the car ?" Usually an
officer will do that in a situation where the officer feels threatened in some way. Wants
to, you know, was worried about something, or the officer's thinking maybe an arrest is
going to take place. And so you ... you can track that too, down the road. That's another
good piece of information to track.
Botchway/ So you said you ... can you go back to the survey portion of that again? You said you
did a survey and what happened?
Barnum/ Yeah, we surveyed the officers and we asked them how they're filling out the box, and
about half of `em said that they're checking the box when the driver says `yes' to the
search, instead of when the officer asks the question. And then the other half s ... were
doing it correctly. But on top of that then ... many of the officers are also asking the
question when they shouldn't be asking the ... or, well I mean I shouldn't say it that way.
Are checking the box when they shouldn't be checking the box at all because they... they
have PC. We only want to track searches that you don't have PC, where the officer really
cannot get into the car. The officer can't get into that car unless that driver says yes.
And then you want to see if that's done disparately. If you have another reason to get in
the car and you're asking `em, it's still... it's not really (laughs) a voluntary search.
You're going to get in the car either way! You're just hedging your bet. And so here's
the hit rates, and it shows this ... this box shows you that ... so this box is the search two
and a half times higher for minority drivers. Hit rates is a ... below 1 number. The way
you convert that is you just ... you invert it, turn it upside down. And it'll give you 1.6.
And so whites, white and Asian drivers, the hit rates were, um, 60% higher for white and
As ... Asian drivers, even though requests were ... 2.54, the odds were ... the times were
2.54 odds higher. So this last table ... is, um ... citation, arrests, search requests, and hit
rates broken out, uh, for each year, and you can see the pattern's pretty clear. For
citations in the first part of the study, white drivers, white and Asian drivers, were more
likely to get a ticket at the end of a stop. Second part of the study, um, minority drivers
were more likely to get a ticket, but the, um ... the odds ratios are particularly high in
either instance, so that's kind of a wash. Tickets don't appear to be any sort of an issue.
If you look at arrests, it's a pretty clear pattern all the way through. Minority drivers are
arrested at higher numbers than non - minority drivers, but again, there can be many
reasons for that. I want to emphasize that. Unless we know the reason for the arrest, we
don't know why that's occurring, and so that is not an indicator of bias in and of itself.
Search requests, which we've talked about in depth. You can look at that. You can see
that that number is, um ... higher than I and so ... that tells you that the, um, requests are
being made of minority drivers, more than non - minority drivers, across all years of the
study. And then hit rates, with the exception of 2006 are always higher for non - minority
drivers than the other way around. Now what's interesting about this, and then I'm going
to shut up finally, is that ... we didn't see the same changes in the data with the outcomes
as we did with the stops. So, that didn't change. It was the stopping behavior more than
the outcome that changed ... due to this, which is interesting. And that is my last slide. I
would be happy to entertain questions and offer recommendations.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 37
Botchway/ Well I guess, I don't know. I kind of wanted to hear the recommendations first.
(loud noises in background) Yeah, I wanted to hear the recommendations first and then
have questions (several talking)
Barnum/ Oh that's fine! Yes! Yes, I'd be happy to do that. So my recommendations are is that
you should continue the study, uh, for at least, uh, a year or two. There's a couple of
reasons for doing this. Um ... I think you'll find, and we've found this when we've done
it in other places, that the officers' behavior will change once their known ... they know
they're being watched, once it becomes public. That's human nature. Um ... as Chi...
Chief Donchez of Davenport said, when the levels of disproportionality went down there,
and I ... I tend to agree with him on this comment, well ... it's probably cause the officers
are being watched, but I don't really care because the numbers are going down. And I
don't know whether that'll happen in Iowa City or not, but it's ... it's, so that's... that's
one thing I would do.
Dobyns/ Was that because of outliers... outlier officers, or was that everybody, regress toward
the mean?
Barnum/ There really wasn't... it really wasn't regression. It was an actual... it wasn't an artifact.
It was an ... it was an actual ... you can watch it go down, and it's...
Dobyns/ But, did the outliers, I mean, was it...
Barnum/ Well, the officers knew they were being watched so...
Dobyns/ Every... so everybody inched toward the mean.
Barnum/ For the most part. There were ... there were still some that were higher, but they have a
tactical unit there too, and theirs had ... pretty significant levels of disparity and uh, those
numbers also came down. Now what's interesting in Davenport is that they have a traffic
unit, and their traffic officers have no disproportionality at all. And if you think about
that, that makes sense because a traffic officer is running radar, you know, they have the
radar gun. They don't see the driver before they decide to stop. Many times, and so they
pull out and when you look at that, uh, we weren't finding any levels of
disproportionality and that really made us feel good about our baseline. We used the
same methodology there. It's a little more complicated in Davenport because of the way,
uh, people ... the areas where they live are more clustered than they are here, but our
methodology agreed 100% with the traffic data. In other words, their traffic officers were
showing the same percentages as what we said their baseline... was, and so, um ... you
know, that made us feel good about the ... the, uh, the strategy there and that's why I'm ... I
feel good about the baseline here too, particularly the idea of it changing. So ... I ... I
think... I think there's a chance, I think there's a good chance that the numbers will
change when the officers know they're being watched. Um, but again, that doesn't nec...
that's not to say that there's a problem now if they do change. The officers, you know I
mean, it's possible that there's no problems and the officers are so cognizant of it that
they're making absolutely sure that there's no problem, even though there was no
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 38
problem in the first place. (mumbled) ...so in other words, for every ... every, uh, person
of color I stop I run out and stop two white guys, you know, an officer may be ... had that
calculus in his or her head. It's possible after it's watched so ... so you have to ... you have
to evaluate that. Secondly, I think that, uh, the data set would benefit if we could, um,
figure out a way to track where the stops are being made, and we've talked about that.
There's... there's some ways of doing that, um, fairly simple ways of doing that, um ... by
laying a grid over the city and assigning little areas, um ... using GPS, and then that's easy
to download into st... stat packages, um, and so that's something that could be done
moving forward. Thirdly, the, uh ... reason for arrest should be tracked. That's... that's an
important piece of information, and then fourth, it's also important, I think, to uh, start
tracking these, uh, search requests more accurately than they're being tracked now, and
try to emphasize to the officers that they can still ask the question but just don't check the
box in the car when the question's asked, when you have another reason to get in there.
Because it's still going to be in the report, and the report's what's going to be in court,
not that little zero or 1 on an Excel spreadsheet. Thousands of `em!
Dobyns/ And the reason due to reasons for arrests, so we can do a sub - analysis more accurately
on non - discretionary... or discretionary reasons for arrests? Is that the reason?
Barnum/ Yes, I would ... it would be...it would be worthwhile to look at the reason why, uh, the
drivers are being arrested.
Dobyns/ Okay.
Barnum/ Now you know there's several ways that could turn out.
Dobyns/ Yeah ... which would be interesting.
Barnum/ Those are my major, uh, recommendations. Um, so if you have any questions then I'd
be happy to entertain them.
Botchway/ So my ... my, uh, my first question speaks to the data that you have now. Is there
ability to dig deeper from ... some of the comments you were making based on, you know,
the amount of time some ... an officer asks a ... an individual to get out of the car, I mean,
do we have that data now to use? So whatever data we do now will be obviously new
data, or whatever... if we decide to do that, it would be new data, but then also "buy" us
data based on the fact that, you know, officers are now aware that we are conducting the
study.
Barnum/ Right.
Botchway/ Okay.
Barnum/ Yes, that ... that's correct. You know, we can't ... we can't go back in time and ... cause
that question wasn't asked, urn ... but it should be tracked in the future and ... and you
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 39
know, you might want to track this a couple years and then down the road, after a period
of time, track it again, much like was done with Louisville and (both talking)
Botchway/ Well, in that case would ... would it be your recommendation to add that particular
portion of it to ... if we decide to move forward, um ... uh, to the study? I mean, you went
through your recommendations but I don't think you had that particular portion of, you
know, whether or not an officer asks an individual to get out of a car, and then other...
other...
Barnum/ Yeah.
Botchway/ ...particular points where bias could be, uh...
Barnum/ Yeah.
Botchway/ ... seen.
Barnum/ There's a couple of other things like that. Uh, you can look at pat -down searches...
Botchway/ Right.
Barnum/ ... for example. Um, you know, yeah, you can track several things. I kind of included
under the umbrella (mumbled)
Botchway/ Okay.
Barnum/ But yes, there's... there's several ways to track. Now ... now again, now all the things
we're talking about are outcome, stop outcomes. So the traffic stop stuff, the whole first
half of my blathering, won't change. It's the second half that we can track a little more
closely to see after a person's stopped if there ... if there's bias.
Botchway/ Going back to the data that we currently have, is there any way to use, you know,
some type of GIS software to track, I mean, it kind of speaks to Michelle's comments,
um, questions earlier, um ... I mean, is that a possibility? I mean, I know it's going to take
a little bit longer, obviously, because you said you had a whole bunch of different
addresses, but can we ... still...
Barnum/ Go back in time you mean?
Botchway/ Yes.
Barnum/ Um, I'm not sure what the data looks like. I've never seen them. I can't...
Payne/ There ... there are companies that will take addresses and GO code them.
Botchway/ Right.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 40
Payne/ So I mean, it costs money but...
Barnum/ Yep.
Payne/ Is it possible, yes?
Barnum/ It is possible if, you know, but I don't know what the data set looks like in terms of the
addresses. I haven't ever seen it, so I ... I ... it would be, uh, unwise of me to answer that
question.
Markus/ What would it accomplish?
Barnum/ You could ... we could go back if we looked at that, um, and look to see where the stops
were made at. And so we could see, for example, if the, um, beat 5 officers were indeed
making stops in that 2A area, and then we would know well that explains that, but it
always begs the question, any time you answer a question it begs a question, and the
question then would be why were they there and not somewhere else? But often there's
answers to those questions.
Botchway/ My next (both talking)
Hayek/ Go ahead, Kingsley (both talking) No, no!
Botchway/ My next question was kind of regarding the Davenport study. As you saw, you said
that, you know, officers were being watched, um, the disproportionality went down.
What was the crime rate, as far as the violent crime or whatever the case may be?
Barnum/ It ... it's funny you ask that question because I was just talking. I ... I give a presentation
there tomorrow in Davenport, and I was talking to the, uh, Chief there and after they
...uh, started looking at their data more closely, they found that ... well, we found that
although their levels of disproportionality were going down, they were still making a lot
more stops in the areas of town with higher minority concentrations, and so that ... that
begs the question too — why are you making all your stops in this area of town and none
out by the mall? And so what they did was started tracking their crime, using GIS and so
forth, and they found that when they overlaid... when they overlaid, and they had ... they
have a higher crime rate than Iowa City does. It's ... it's significantly higher there. But
they found that ... when they tracked their shots fired calls in Davenport, it pretty much
followed where they were making their traffic stops. And ... it did, their tactics did tend to
lower crime, although that's preliminary, okay, so ... I ... I don't, I can't say that for
certain. We're going to look at that as we move forward, but the question is ... is it ... is it
really lowering the crime rate or just moving it? So, if you're in one neighborhood... do
(noise on mic) Oops! (laughs) Do the people all move three blocks over? It's called,
uh, displacement. And so that's some issues that we're going to look at. But, um, it
seems that what they are doing in the ... is, um, is lowering crime, but it's again, very,
very preliminary.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 41
Payne/ I ... I find it interesting that if you look at the chart on page 29 with the estimate of violent
crime rates, that there's a little tail going back up right now, from 2010 to 2012, and at
the same time we reduced the beat cops in 2A, and I would like to know if there's a
correlation there.
Barnum/ That's possible.
Payne/ But you have to know where the violent crime was, you know, on the map to ... to know if
that has any correlation. I just think that that's interesting.
Barnum/ You can kind of...you can kind of see where it's at by looking at whether it's in
Wetherby or Grant Wood and ... and you know, so you can kind of tell ... that way. But
you know it's ... it's, um, it's more ... these are just ... these are actual offenses in these
graphs. In Davenport what they're tracking are calls. Calls into the police department,
which is a little more revealing. You know, someone calls up and someone just fired a
gun. That happens a lot up in Cedar Rapids too. You get that call a lot up there, and
when the officers get there, you don't find anything. So no report is ever generated. It
doesn't go to the FBI. It doesn't end up in the statistics, but it still puts an officer in the
area. So there's those ... those sorts of things.
Throgmorton/ So ... so I'm thinking that maybe ... I'm going to make a suggestion here and
I ... I'm not sure who might or might not agree. It could well be that what happened here
in 2008 and 9, or thereabouts, was, um something of an overreaction to a perceived
problem. Uh, to a perceived increase in violent crime. Now, I know there were violent
incidents, and I do not want to understate them. Uh, but the question is, the ... the, uh, wh
...whether, um, the response was appropriate, and ... and you know, I know people have
differing views about this. But it matters because if we look at, uh, the ... the current
events that have recently taken place in the Pheasant Ridge area, then one sees something
also worrisome because there have been shots fired, with three incidents as far as I know,
Sam. So we want to ... we want to deal with that. We don't want people walking around
shooting other people and so on. Uh, but we don't want to overreact either. And ... and
suddenly start tracking basically black people because some black people did the
shooting. Uh, it ... it's not that simple.
Dobyns/ Jim, we're tracking geographic areas. We're not necessarily tracking ethnicity.
I ... think you're getting a little bit far off there.
Throgmorton/ No I ... I don't fol ... I don't understand what you mean about that. This whole
discussion (both talking)
Dobyns/ ... geography area. It isn't an ethnic area necessarily. I don't think you can ... I think
you're making some assumptions that aren't entirely relevant to tonight's discussion.
Throgmorton/ Uh ... well ... I ... I doubt that, but ... but I'll just make my point and leave it at that.
Uh, the ... when we look ahead, I would think we would want to make sure that our
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 42
response is appropriate to the actual events and focused on the people committing the
offenses.
Barnum/ I will say this much (clears throat) I think that the, uh, Police Department has been
extremely open in providing their data and ... and allowing this to be tracked, which
doesn't occur everywhere. It really doesn't, and so they're willing to put themselves out
there and look at this, which leaves you open to judgments. Um, it's a very difficult
situation for a police administrator when there's ... when there are instances like this. It's
sort of a damned if you do, damned if you don't. If you don't address the crime problem,
then there's, you know, ramifications there, as well. And so I think that that's a ... that's
an issue that, you know...it...it ultimately is up to the community to decide what's an
overreaction and what's not in terms of police response. It's really not anything I can
give you in numbers.
Markus/ Chris, why ... why do you only recommend two years? Of going out.
Barnum/ Well, I ... I don't necessarily ... I would do it at a minimum that much. You know, this
might be something you might want to consider doing for a while, you know. That's the
attitude Davenport has. They're just... they've... they've agreed, um ... to do it, and it's...
it's really amazing watching the situation in Davenport, because there was a ... there was a
fair amount of tension between the advisory committee, which was essentially the civil
rights, uh, commission and some other people and the police department at the beginning
of the study, low levels of trust, but as this went on and the process was open, it...it
significantly increased the trust level, and I think all stakeholders there are quite pleased
that the ... it's moving forward.
Markus/ So you indicated that the study has seemingly driven down the DMC, the annual
reviews. How long has it gone on?
Barnum/ They've do it ... they've done it now three years. They do it ... they've done it, uh, I've
reported three times there. And tomorrow (both talking)
Markus/ So are they going to continue doing it, annually?
Barnum/ As far as I know they are, yeah. They've... they've indicated they are (laughs) It's
always subject to funding, of course, but they ... both parties want to do it. I ... I don't...
imagine that they'll stop. Both the, uh, the advisory commission, which is essentially the
civil rights commission and other groups, and the police department, want to do it.
Markus/ It just seems like a much move factual way of, you know, analyzing the information and
dealing with the issue of DMC.
Barnum/ I agree. I ... I think, you know ... I think by putting it out there, and everybody looking at
it, that's the best way to handle it. I think the worst thing that could be done is, well, the
worst thing is not to collect the data, and the second worst thing is not to look at the data.
I don't know which is worse actually.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 43
Payne/ So your recommendation is to continue doing it to a point in time when we don't think
we need to do it anymore?
Barnum/ Well I... see, I don't know what to tell you there. I mean, I can't tell you do it (both
talking)
Payne/ Well we would have to make the decision (both talking) when that point in time is...
Barnum/ Right.
Payne/ ...but obviously you're not up here looking for us to say, yeah, we're going to do this for
the next 25 years.
Barnum/ Right. No.
Payne/ You're... you're just saying continue doing it, because it's...
Barnum/ No I would ... my ... my recommendation is I would continue to look at this. I think if
you stop now you would raise a lot of questions.
Markus / And can't you get to a point though that the data collection is simplified over time.
Barnum/ Yes. It would become simplified.
Markus/ The analysis is still complex.
Barnum/ Yeah, uh, the thing that we'll have to continually do is update the baseline. Make sure
that's not changing. That's the main thing to look at.
Hayek/ So now that we have Dr. Barnum's, uh ... uh, data report and presentation, you know, if
we shift to staff, what ... what's the next move on ... on our end? Do you look at that,
Chief, or Tom?
Hargadine/ Our goal is to reduce the disproportionate minority contact numbers. Period. My
staff and I remain committed to vigilance and will take seriously all complaints regarding
DMC, or any unprofessional conduct. The release of data is an opportunity for the
Department to grow and I outwardly express our commitment to build relationships and
protect all in the community with the same high standards and professionalism. Uh, I ... I
concur with Dr. Barnum. I would like to see this go forward, as well. Um, we've gotten
a lot of, uh, hard work out of, and I'd like to commend Dr. Barnum and his staff, as well.
Um (clears throat) you know, originally when he and I first met, I ... I wasn't intending for
this to go public. Um, I was a ... to be used as an internal tool. I didn't think I had a
problem at the time, but I wanted to know if I did or not, and that's why we ... we did this
as long ago as we did. As we went through the DMC process, and I knew what, uh, the
questions that were coming in, uh, from a couple of the, uh, Ad Hoc, uh... Committee
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 44
members, uh, that's when I relayed to Tom that we were undergoing this study, and that
we've had this long -term relationship with Dr. Barnum. So ... at that point we decided it
was time to go public with it, and release the data.
Markus/ And to do I think more analysis on the data itself. So...
Hargadine/ Right.
Markus / And that's what's been completed here recently by Dr. Barnum.
Botchway/ So, um, so this is kind of a question for both. I mean, I guess a lot of the comments
that came out of the Ad Hoc Diversity Committee were yes attributed to traffic stops.
Um, whether or not, you know, some ... an individual was stopped too many times or
whether or not they felt that stop was, urn ... uh, a good stop, whatever the case may be,
my question I guess, since you're recommending that we move forward, I'm actually,
you know, obviously, um, as the doctor said, more interested in some of the, um, the bias
elements that you can't necessarily tell from this data because , you know, to make kind
of a, I guess a broad... generalization of what we've kind of looked at is that you're
saying that disproportionality exists; however, you're not sure what you can attribute it
to, except for the fact that there were ... there was a focus in particular areas of this
community, um, where there was a higher number of minorities. Um, from that
obviously, you know, I understand your recommendation to move forward, but again, a
lot of the complaints and a lot of the things that were, whether anecdotal stories or, um,
compla ... I mean, complaints, um, documented to you was more about the nature of the
actual stop. And so whether or not, uh, this individual treated me in a fair manner, or
whether or not, um... when I say the individual I mean the officer. Whether or not the
officer treated me in a fair manner. So my question I guess, going forward, is, is there
going to be, uh .... more analysis as far as that is concerned. I mean, off the top of my
head while we were talking about it, I didn't know whether or not in your, uh ... your
experience you've seen like a suggestion box, you know, cause usually a traffic stop
takes some time to actually take place. I mean, you get stopped; the officer comes asks
for identification information; and so there's some time where an individual sitting there
that presumably they could fill out some type of suggestion box or ... complaint box,
whatever... for whatever reason, that you could get information from the actual individual
that's in the stop, that you wouldn't necessarily get down the road if you're asking that
individual to come to the police office and, you know, file a complaint or go wherever
and file a complaint. Is that something that ... you and you would recommend, or, I mean,
are we going to ... look at things I agree ... look at the issue as far as disproportionality's
concerned but don't necessarily speak to all the issues that were raised by the Ad Hoc
Diversity Committee, and other people that came to the forums, as far as, again, how they
were treated when the ... when the stops occurred.
Hargadine/ One of the things that, you know, outside of the data we ... we video every car stop.
When ... when they turn the lights on, the camera starts up. And it's audio, uh, recorded,
as well. So those, when we have a specific complaint, when we know of a particular
instance, um ... we ... we have the ability to go back and pull up that particular traffic stop
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 45
and listen to what transpired. So ... when you're talking about the specific case, we can
look at those specific cases, and we do. We also randomly, uh, pull up. Each supervisor
has to randomly look at a number of different videos per month on ... on officers, and then
they just look at their techniques. They're also looking for safety issues. Is this ... is this
officer have, uh, any training ... I mean, is he doing it in the most tactically safe manner.
So those are ... they're looking at it from a couple of different, uh, eyes, uh, but they're
also looking at politeness, you know, independent of race. Are they, is this person being
as professional as they possibly can. Um, we also do it, um ... even more routinely on
our ... on our newer officers. So, um, when someone's in their first year, they've been
released from field training. It's very common that we're... they're watching those...
those video tapes. Um, for technique, courtesy, and professionalism. (both talking) You
had a long question there! So (both talking)
Botchway/ ... can that information be, you know, used in the new data that we're going to get,
you know, later on down the road? Because I ... I believe that's important information for
us to have, I mean, as you said, you know, having ... if an officer asked an individual to
get out of the car, that sometimes means there's bias that's indicated. And so you know
whether or not a particular officer, and granted, we don't need to know the officer's
name, but whether or not you're tracking that officer from point A to point B over a
period of time, and um, they are being less polite, um, to one or ... to a specific group,
compared to another group, I mean, that's information that I would want to know, from a
disproportionality standpoint.
Payne/ But isn't professionalism and ... respect and ... what you're talking about a matter of
opinion? I mean ... you could think that somebody's disrespectful, and I may not think
they are. Or professional and I may not think they are, I mean, it's so subjective
that ... how can you ... how can you track that and have any real meaning to it?
Botchway/ Well that's why I don't plan on (laughter)
Payne/ I don't mean you! I mean anyone!
Botchway/ Well I mean that's where we want the Chief and the Dr. to work on some type of way
(both talking)
Hargadine/ No, you bring up a good point. We do have officers that, uh, especially those that are
younger and those that are doing a lot of stops, they tend to bring in more complaints, of
all kinds, than ... and I'm not talking about just racially biased complaints, but you know,
I got pulled over, uh, and he was rude to me. We get ... we get those. So, uh, if...when
we have officers that have a higher number than ... than average, then yeah, we ... we start
looking at those video tapes and ... and frequently there'll be a counseling session going
on and ... and they sit down with a supervisor in front of the screen and say, `Take a look
at this and tell me is there a better way that we could be doing this.' And they generally
critique themselves. So that does go on. Does it come to a report that ... you know, for
you guys? Uh, that ... that kind of coaching. No, but it does go on.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 46
Dobyns/ Professor, is there any metric that you can think of that... Kingsley might be ... might be
help that?
Barnum/ One problem I think with using a suggestion box is that, uh, at the time when it's being
filled out there's... emotions are generally running high. You know, you... and if you're
turning that thing in after you receive the ticket, you know, um... so it could be biased
that way. One ... one way this could be done, it's done other places, is for outside judges
to look at video. Um, so a team of researchers, blinded to what's going on. You'd
randomly select video, you know, just randomly. They'd watch it and they'd ... you could
develop a checklist.
Dickens/ Is that something you would do?
Barnum/ Well that could be done, yes.
Dickens/ As part of the...
Barnum/ Yeah, that ... that could be done.
Dobyns/ For a cost. (laughter)
Barnum/ Well I mean (several talking) I mean... something like that is a little bit more labor
intensive, and you know, but you know, with grad... graduate students, it's ... (laughter
and several talking) So something like that is possible.
Dobyns/ If we could insert that into future data gathering, at least maybe for a sub - portion,
um ... I think that would be reasonable. To consider, because we're ... not the whole thing
but if we could take a look at some of the data and decide if it's garbage, I guess, or not
(mumbled)
Hargadine/ We have talked about the coordinates and the GPS issues, and that's something we're
going to be looking at, um, at future data sets.
Hayek/ So we'll ... but I ... I just want to make sure we transition from this report to what ... what
we as an operation do going forward, and we've got a set of recommendations. You
know, will there be a point at which, you know, staff reports back to us and says this
is ... this is our plan, this is what we're doing moving forward with the continued
de ... collection of data, the more refined collection of data, and, you know, changes here
and there, and so on and so forth.
Markus/ Yeah, we'll report back to you. I'm not sure that it ... it's going to require Council
approval, uh, just because of the dollar amounts and I don't want to encourage Council
approval so that Dr. Barnum changes the rates that he's been charging us so far (laughter)
which have been pretty reasonable, by the way. Um, so ... I think we're, you know, what
I'm hearing is there's support to continue this process. (several talking)
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.
June 16, 2014 Iowa City City Council Special Work Session Page 47
Hayek/ Without a doubt!
Markus/ I think we need to adjust some of the data points, uh, and we need to instruct ... get into a
full instruction in the officers and how they're reporting things and make sure those
things are all occurring... correctly.
Hayek/ Okay.
Markus/ But ... I'm hearing...
Hayek/ I don't think anyone (several talking) is concerned... yeah.
Markus/ And if there's anything that trips the ... you know, the wire that it has to come back for
Council approval, we'll bring it back, but otherwise we'll report back what we're doing.
Mims/ Yeah, I think having a report back from staff, and having it at a formal meeting would be
really good from a publicity standpoint, because so many people don't necessarily pick
up on what we do in work session.
Markus/ Well this is being taped tonight. This will be shown as well.
Hayek/ Okay.
Throgmorton/ Yeah, uh, I think I'd like to say two brief things. One is it ... I ... I want to make
sure I pay attention to the fact that there's interpretation ... interpretation involved in..in
this, so we can't ... nobody can just look at the data and say this is what the data say. So,
there's always going to be some interpretation involved. But the other thing, I ... I
personally want to do is thank Sam for initiating this study, uh, and for presenting it to us.
I think, and Chris, I think you've done a superb job of, uh, working our way through it.
You've helped me understand the written report better than I did before. So thanks to
both of you and uh, I'm sure, Sam, you're going to follow up on this, and I have total
confidence in that.
Botchway/ I did have a quick thing. After the meeting can I talk to you about the back pages?
The appendix? Okay. (several talking and laughing)
Hayek/ Okay, I want to thank, uh, Chief Hargadine and the Police Department for what they do
for the City every day of the year and for, uh, their role in ... in this data, and Dr. Barnum,
you and your ... your team, excellent work, and everyone else involved at the staff level, I
appreciate that as well. So I think this was a very good session. Uh, let's uh, adjourn at
this time and we'll see you tomorrow, uh, at 5:00. Thank you.
This represents only a reasonably accurate transcription of the Iowa City City Council special
work session of June 16, 2014.