Download presentation
Presentation is loading. Please wait.
1
Dubuque PD Traffic Stop Analysis
June 2017
2
Goal of the Study Look for Disproportionality in:
(1) The decision to make a stop, and (2) The outcome of the stop
3
Definitions Disproportionality Stop Disproportionality Outcome
A difference between police data and an expected value Stop Disproportionality A difference between racial demographics of police stops & a benchmark Outcome A difference in the racial demographics percentages of stop outcomes
4
Decision to stop We analyzed traffic stops made by the DPD between January 1st, and December 31st 2015. Goal: to compare the percentage of minority drivers stopped to a valid benchmark
5
Where the traffic stops were made
8
Traffic Observer Locations and Findings
9
Observation Locations and Numbers
Correlation r for inter-surveyor observations within each location equals 0.63
10
Day verses Night Observations
Night and day benchmark values are not significantly different from one another (mean difference = 0.001; t = 0.057; p = 0.95)
11
Correspondence between bench & census
Census and Max benchmark values are not significantly different from one another (mean difference = 0.01; t = 0.47; p = 0.64)
13
Zone 1 adjustments Break into block groups
14
Final Benchmark Values in Observation Zones
Census and Final benchmark values are not significantly different from one another (mean difference = 0.032; t = 0.995; p = 0.33)
15
DPD stops compared to the benchmark
16
Stops compared to benchmark
17
Weighted average across all zones for department as a whole
A Low Level of Disproportionality Estimated margin of error is greater than
18
Days vs Nights Days zone min % bench stops 1.1 0.12 0.11 170 1.2 0.08
219 1.3 0.17 0.15 393 4 0.02 0.05 214 5 0.24 0.2 330 6 101 7.1 0.23 64 7.2 0.1 106 9 0.10 176 11 0.03 349 99 0.07 1024 Nights zone min % bench stops 1.1 0.15 0.11 142 1.2 0.17 101 1.3 0.20 240 4 0.06 0.05 52 5 0.23 0.2 350 6 0.19 0.12 73 7.1 0.28 39 7.2 0.14 0.1 37 9 0.08 170 11 0.07 174 99 0.09 660
19
Disparity Index Summary
Days Nights Dept. Disparity 0.009 0.045 0.024 Stops 3146 2038 5184
20
Discussion (1) The mean level of disproportionality for the department as a whole is low at roughly 2 percentage points above benchmark values. Within margin of error (2) The mean level of disproportionality for stops occurring during daylight hours is less than 1 percentage point above benchmark values. Within margin of error (3) The mean level of disproportionality for stops occurring during nighttime hours equals roughly 4.5 percentage points above benchmark values. Just outside margin of error
21
Stop Decision Conclusions
Analyses found no significant levels of disproportionality for the department as whole. Specifically, Virtually no disproportionality in traffic stops during daylight hours (60% of all stops occur in daylight). Low levels of disproportionality during nighttime hours.
22
Citations, Arrests, and Search Requests
Outcomes Citations, Arrests, and Search Requests
23
Citations Odds Ratio = 1.21 Minority drivers are slightly more likely to be ticketed on a traffic stop
24
Arrests (raw) Odds Ratio = 3.49
Minority drivers are more likely to be arrested on a traffic stop
25
Arrests (adjusted for warrants)
Odds Ratio = 2.66 Minority drivers are more likely to be arrested on a traffic stop
26
Search Requests Odds Ratio = 6.08
Note: The number of search requests is very small. Therefore, the odds ratio may not be a valid indicator of disproportionality. The data show that the DPD request to search very infrequently. On average this occurs once every 260 stops.
27
Moving or Equipment Violation
Reason for the Stop Moving or Equipment Violation
28
Moving Violations Odds Ratio = 0.69 (1.44)
Whites are more likely to be stopped for moving violations
29
Equipment & Registration Violations
Odds Ratio = 1.45 Minority drivers are more likely to be stopped for an equipment violation
30
Multivariate Analyses
31
The control variables used
Time of day Gender Type of violation Age of driver Location of the stop
32
Logistic Regression Results
The results of logistic regression show that controlling for important variables does not substantively change the results for citations, arrests and search requests. However, the results also show that racial differences in stops for moving & equipment violations become no longer significant
33
Conclusions Very low levels of disproportionality in traffic stops. Minority member drivers are not over represented in stops. No evidence of racial profiling Low levels of disproportionality in citations. White drivers and minority members drivers are nearly equally likely to receive a citation as the result of a stop Higher levels of disproportionality in arrests. The odds that minority member drivers are arrested on a traffic stop are about two-and-a-half times higher than for white drivers.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.