2023-24-CTRP3-Traffic-Stop-Analysis-and-Findings-Report

www.ctrp3.org
TRAFFIC STOP 
DATA ANALYSIS 
AND FINDINGS, 
2023-24
RELEASED 
JUNE 2026

 
 
 AUTHORS 
 
Dr. Matthew B. Ross, Ph.D.  
Associate Professor 
School of Public Policy & Urban Affairs and Department of Economics 
Northeastern University 
 
Ken Barone 
Associate Director 
Institute for Municipal and Regional Policy 
University of Connecticut 
 
CONTRIBUTORS 
 
James Fazzalaro 
Senior Research and Policy Analyst 
Institute for Municipal and Regional Policy 
University of Connecticut 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
TABLE OF CONTENTS 
Preamble .................................................................................................................................................................................................. ii 
Executive Summary of Findings ...................................................................................................................................................... i 
E.1: Summary of Methods .............................................................................................................................................................. i 
E.2: Summary of Findings ............................................................................................................................................................. ii 
Highlights from the Analysis ....................................................................................................................................................... ii 
Background ............................................................................................................................................................................................. 1 
I: Methodological Approach Underlying the Analysis ........................................................................................................... 2 
II: Characteristics of Traffic Stop Data ......................................................................................................................................... 4 
III: Analysis of Traffic Stops, Solar Visibility ........................................................................................................................... 12 
IV: Analysis of Traffic Stops, Synthetic Control ..................................................................................................................... 18 
V. Analysis of Post-Stop Enforcement Actions ....................................................................................................................... 21 
VI: Analysis of Vehicular Searches .............................................................................................................................................. 27 
VII: Follow-Up Analysis ................................................................................................................................................................... 31 
VII.A: Meriden Population Breakdown by Race/Ethnicity ..................................................................................... 31 
VII.B: Traffic Stop Breakdown by Roadway and Race/Ethnicity ......................................................................... 34 
VII.C: Review of Traffic Stops Along Selected Roadways ......................................................................................... 37 
VII.D: Traffic Stop Distribution for Meriden Officers ................................................................................................ 41 
VII.E: Post-Stop Outcome Review ...................................................................................................................................... 42 
VII.F: Additional Contributing Factors ............................................................................................................................ 47 
VII.G: Summary of Findings and Conclusion ................................................................................................................. 48 
References............................................................................................................................................................................................. 50 
Glossary of Terms .............................................................................................................................................................................. 52 
 
 

 
 
PREAMBLE 
This preamble was written by an ad hoc committee of the Connecticut Racial Profiling Prohibition 
Project advisory board and unanimously endorsed by the board on December 6, 2018.  
1. Racial Profiling has historically occurred and continues to occur throughout America. 
2. The Alvin W. Penn Racial Profiling Law enacted by the Connecticut General Assembly in 1999 
required state and local police to collect traffic stop data and report the data to the state. 
3. The 2011 federal investigation into the East Haven Police Department brought this issue to 
the forefront in Connecticut again and led to the Connecticut General Assembly updating the 
Profiling Legislation in 2012.  
4. Disparities across racial and ethnic groups occur in traffic stops in Connecticut. 
5. Enforcing the law’s data reporting requirement and collecting and analyzing racial disparities 
in traffic stop records in the primary charge of the advisory board. 
a. A broader analysis utilizing multiple methodologies in the preferred method for 
measuring the presence of racial disparities in traffic enforcement; 
b. Although no measure is 100% accurate in measuring disparities, the analysis utilized 
in Connecticut is sufficient in determining the presence of disparities; 
c. We will continue to modify and refine our methodologies based on the best available 
research and accepted practices in the field. 
6. We will take a proactive approach to understanding, explaining, and addressing disparities 
found in the analysis by: 
a. Utilizing input from all stakeholders to understand the underlying causes for such 
disparities; 
b. Clearly explaining to the public and stakeholders if there are justifiable reasons for 
such disparities;  
c. Reporting to the Office of Policy and Management instances where the Connecticut 
Racial Profiling Prohibition Project Advisory Board believes that a police department 
is in violation of the Alvin W. Penn law.  
 
 

i 
 
EXECUTIVE SUMMARY OF FINDINGS 
This is the tenth analysis conducted by the Institute for Municipal and Regional Policy (IMRP) at the 
University of Connecticut (UConn)1. The IMRP and Connecticut Racial Profiling Prohibition Project 
advisory board is particularly well-known for developing the technical framework of the 
"Connecticut Model," a pioneering approach designed to identify and mitigate racial and ethnic 
disparities in police traffic stops. Our approach has been adopted by multiple states, endorsed by 
advocacy organizations, and is nationally recognized as the gold standard approach for analyzing 
traffic stop data for evidence of disparate treatment.  
This analysis examines reported traffic stop data from the 2023 and 2024 calendar years, including 
a three-year aggregate assessment of stops across the study periods. Data were submitted by 95 
municipal police departments, the Connecticut State Police, and several special police agencies. 
In total, Connecticut law enforcement agencies conducted approximately 358,000 traffic stops in 
2023 and 413,000 in 2024. White non-Hispanic drivers represented approximately 58% of all stops 
in both years, while Black drivers accounted for 18–19% and Hispanic drivers for 19–20%. 
Traffic enforcement activity increased by 14% in 2023 relative to the previous year and rose by an 
additional 15% in 2024. Of all traffic stops conducted during the study period, approximately 75% 
were initiated by municipal police departments, while the Connecticut State Police accounted for the 
remaining 25%. 
E.1: SUMMARY OF METHODS 
For the past two decades, analyzing racial disparities in policing data has been a critical policy tool 
for evaluating the potential presence of racial and ethnic bias within various jurisdictions. This report 
presents a statistical assessment of policing data for Connecticut to provide a clear, transparent, and 
unbiased evaluation. The report is structured to guide the reader through several analytical tests, 
each differing in assumptions and levels of scrutiny.  
• 
Solar Visibility Analysis: Solar visibility analysis compares the rate at which White and 
non-White drivers are stopped during daylight to the rate at which they are stopped in 
darkness when it is harder for the officer to observe the driver’s race.  When there is a 
higher relative rate of non-White drivers stopped in daylight than in darkness, it indicates 
racial bias. This method is among the most rigorous approaches available for evaluating 
racial disparities in Connecticut’s traffic stop data. 
• 
Synthetic Control Analysis: This method compares traffic stop data from individual police 
departments to a synthetic benchmark based on stops in other departments with similar 
characteristics. The goal is to assess whether non-White motorists were disproportionately 
stopped in specific jurisdictions relative to expected stop rates. 
• 
Post-Stop Enforcement Action Analysis: This method examines each traffic stop 
conducted and then compares the outcomes of the stop between White and non-White 
drivers. Outcomes can include arrests and other discretionary law enforcement actions 
(searches, tickets, warnings, amount of time stopped). When there is a different rate of a 
 
1 The Institute for Municipal and Regional Policy was previously located at Central Connecticut State University. 

ii 
 
specific outcome for non-White drivers compared to White drivers who were stopped 
under similar circumstances, it can indicate racial bias. 
• 
Search Hit Rate Analysis: This method examines each traffic stop where a search is 
conducted and then compares the rates of contraband found between White and non-White 
drivers. “Contraband” is an illegal item, such as drugs, weapons, and stolen property. When 
there is a lower rate of contraband found for non-White drivers compared to White drivers 
who were stopped under similar circumstances, it can indicate racial bias. 
We use this multi-test approach to safeguard against potential errors, reducing the possibility of (1) 
false positives- where a disparity is detected where none exists, and (2) false negatives- where a real 
disparity goes undetected. Each method has inherent drawbacks based on the volume and structure 
of the data available for this analysis. However, if we find consistent disparities across Connecticut 
or within specific police departments, it indicates an area for researchers to investigate further to 
determine if the disparities result from specific policing practices that can be changed. 
E.2: SUMMARY OF FINDINGS  
The Solar Visibility Analysis, among the most rigorous approaches available, shows a modest upward 
trend in the likelihood of Black motorists being stopped during daylight hours. However, most results 
were not statistically significant across the years examined, including 2024. For Hispanic motorists, 
a similar upward trend was observed. While results were sporadic in 2021 and 2022, by 2024, nearly 
all findings were statistically significant, indicating a measurable disparity for Hispanic drivers. 
When comparing outcomes of traffic stops, such as arrests, tickets, and stop duration, we found Black 
and Hispanic motorists were more likely to experience a more severe outcome, an arrest, or a 
precursor to an arrest, even after controlling for circumstantial factors. Our evaluation of search data 
also revealed that Black and Hispanic motorists were disproportionately subjected to a search and 
less likely to be found with contraband.  
Highlights from the Analysis  
Solar Visibility Analysis: 
Compares the rate at which White and non-White drivers are stopped during daylight to the rate at 
which they are stopped in darkness, when it is harder for the officer to observe the driver’s race. When 
there is a higher relative rate of non-White drivers stopped in daylight than in darkness, it indicates 
racial bias.  
Statewide Estimates 
o Black motorists: Occasional statistically significant increases, but the effects were 
small and not consistent across robustness checks.  
o Hispanic motorists: Statistically significant increase ranging from 2% to 6% in 2024. 
The disparity was modest in size and indicates that Hispanic motorists are more likely 
to be stopped during daylight, when their race is more discernible.  
Agency Estimates 
• 
2023: 

iii 
 
o Black motorists: 2 State Police Troops and 3 municipal agencies were identified with 
statistically significant disparities. In these agencies, Black motorists are more 
frequently stopped in daylight compared to darkness.  
o Hispanic motorists: 1 State Police Troop and 1 municipal agency were identified with 
statistically significant disparities. In these agencies, Hispanic motorists are more 
frequently stopped in daylight compared to darkness.  
• 
2024: 
o No agency was identified with a statistically significant disparity for Black and 
Hispanic motorists alone. However, when the demographic groups were combined, 
there was 1 State Police Troop and 3 municipal agencies identified with a statistically 
significant disparity.  
Synthetic Control Analysis: 
Compares traffic stop data from individual police departments to a synthetic benchmark based on stops 
in other departments with similar characteristics.  
Agency Estimates 
• 
2023: 
o 
Black motorists: 11 municipal agencies were identified with statistically significant 
disparities. In these agencies, Black motorists are more frequently stopped by police 
relative to their respective benchmark group.  
o 
Hispanic motorists: 8 municipal agencies were identified with statistically 
significant disparities. In these agencies, Hispanic motorists are more frequently 
stopped by police relative to their respective benchmark group.  
• 
2024:  
o 
Black motorists: 7 municipal agencies were identified with statistically significant 
disparities. In these agencies, Black motorists are more frequently stopped by police 
relative to their respective benchmark group.  
o 
Hispanic motorists: 5 municipal agencies were identified with statistically 
significant disparities. In these agencies, Hispanic motorists are more frequently 
stopped by police relative to their respective benchmark group.  
Post-Stop Enforcement Action Analysis: 
Investigated racial disparities in post-stop outcomes such as arrests and other discretionary 
enforcement actions (arrests, warnings, stop duration, non-safety related enforcement). 
 
Statewide Estimates 
o In 2023 and 2024, Black and Hispanic motorists were significantly more likely to 
experience a more severe outcome, an arrest, or a precursor to an arrest, even after 
controlling for circumstantial factors.  
o Black motorists were 13.5% (3.5 percentage points) more likely to experience a more 
severe outcome in 2023, and that increased to 15.5% (3.8 percentage points) in 2024.  
o Hispanic motorists were 20.9% (5.3 percentage points) more likely to experience a 
more severe outcome in 2023, and that increased to 23.1% (6 percentage points) in 
2024.  
 

iv 
 
Agency Estimates 
• 
2023: 
o 
Black motorists: no agencies were identified.   
o 
Hispanic motorists: 1 agency was identified with statistically significant disparities. 
In this agency, Hispanic motorists experience more severe stop outcomes. 
• 
2024:  
o 
Black motorists: 1 state police troop was identified with statistically significant 
disparities. In this agency, Black motorists experience more severe stop outcomes. 
o 
Hispanic motorists: 1 agency was identified with statistically significant disparities. 
In this agency, Hispanic motorists experience more severe stop outcomes. 
Search Hit-Rate Analysis: 
Examined racial disparities in the likelihood of a discretionary search resulting in evidence being found. 
 
Statewide Estimates 
o In 2023 and 2024, Black and Hispanic motorists were disproportionately subjected 
to a search and were less likely to be found with contraband.  
Agency Estimates 
• 
2023: 
o 
Black motorists: 3 municipal agencies were identified where police conducted 
searches of Black individuals at disproportionately higher rates relative to their 
success in finding contraband. 
o 
Hispanic motorists: 1 state police troop was identified where police conducted 
searches of Hispanic individuals at disproportionately higher rates relative to their 
success in finding contraband. 
• 
2024:  
o 
Black motorists: 2 municipal agencies and one state police troop were identified 
where police conducted searches of Black individuals at disproportionately higher 
rates relative to their success in finding contraband. 
o 
Hispanic motorists: 1 municipal police agency was identified where police 
conducted searches of Hispanic individuals at disproportionately higher rates 
relative to their success in finding contraband. 
Each agency or troop could be identified up to eight times across four analytical tests and two years. 
Because each test varies in statistical rigor, results were weighted accordingly, with greater emphasis 
placed on the Solar Visibility Analysis due to its higher precision and relevance to this research. These 
findings should be interpreted in totality, as focusing too heavily on any single test or year would not 
provide an accurate assessment. 
Overall, 38 municipal police agencies and 4 State Police troops were identified in at least one test in 
one year. However, only 10 municipal agencies and 2 State Police troops were identified across at 
least one measure in both 2023 and 2024. Among these, only one municipal agency demonstrated 
statistically significant disparities for Black or Hispanic drivers across multiple tests. Based on our 
evaluation and prior experience, we determined that only the Meriden Police Department warranted 
further analysis. A summary of the detailed analysis can be found in Section VII of this report.  

1 
 
BACKGROUND  
Connecticut’s Alvin W. Penn Racial Profiling Prohibition Act (Public Act 99-198), enacted in 1999, 
prohibits law enforcement officers from stopping, detaining, or searching a motorist  on the basis, in 
whole or in part, of the perceived racial or ethnic status of such individual, except when such status 
is used in combination with other information when seeking to apprehend a specific suspect whose 
racial or ethnic status is part of the description of the suspect (Connecticut General Statutes §§ 54-1l 
and 54-1m). 
In 2012 and 2013, the Connecticut General Assembly strengthened the law through Public Acts 12-
74 and 13-75, establishing a statewide system to address racial profiling concerns. These 
amendments created the Racial Profiling Prohibition Project Advisory Board to advise the Office of 
Policy and Management (OPM) on standardized methods for collecting and analyzing traffic stop 
data. The Institute for Municipal and Regional Policy (IMRP) at the University of Connecticut was 
designated to oversee the design, evaluation, and management of the state’s racial profiling study. 
Beginning October 1, 2013, police agencies were required to submit traffic stop data electronically to 
the state’s Criminal Justice Information System (CJIS) using the new OPM-approved methods. The 
amended law also authorized OPM to impose penalties, including the withholding of state funds, for 
noncompliance by municipal police departments, the Department of Emergency Services and Public 
Protection (DESPP), or other law enforcement entities. 
The project received support from the National Highway Traffic Safety Administration (NHTSA) 
through a grant administered by the Connecticut Department of Transportation (CTDOT). Since 
2012, the Advisory Board, IMRP staff, and working groups—focused on Data, Community Outreach, 
and Policy—have collaborated with law enforcement and community stakeholders to ensure 
effective and consistent data collection across agencies with varying technological capacities. 
The IMRP maintains the project website (www.ctrp3.org), which provides public access to meeting 
materials, press releases, and statewide updates. In partnership with the Connecticut Data 
Collaborative, all traffic stop data are published quarterly, allowing the public to download raw data, 
view summary tables, and use interactive tools for independent analysis. 
In addition to improving data systems, the project emphasizes training, transparency, and 
community engagement. Initiatives include public education campaigns, the development of a fair 
and impartial policing training program, and a continuing series of public forums designed to 
strengthen communication and trust between law enforcement and the communities they serve. 
 
 

2 
 
I: METHODOLOGICAL APPROACH UNDERLYING THE 
ANALYSIS 
Assessing racial disparities in policing data has been a crucial policy tool for the past two decades, 
helping to evaluate whether racial bias exists within a given jurisdiction. Although public support for 
the fair treatment of all races and ethnicities has long been widespread, recent national headlines 
have intensified this issue, sparking a heated debate over policing policy. The statistical evaluation of 
traffic stops by police in Connecticut represents a significant step toward fostering transparent, data-
driven dialogue between law enforcement and the public. The goal of this report is to present the 
results of that evaluation in a clear and unbiased manner. 
This statistical analysis is guided by three key principles, which form the foundation of the research 
process and inform the selection of results to be shared with the public. Understanding these 
principles is crucial to interpreting the technical portions of the analysis. Presenting them at the 
outset gives readers the necessary context to understand the overall approach. 
• 
Principle 1: Recognize that statistical evaluations can identify racial and ethnic disparities 
indicative of racial/ethnic bias and potentially discrimination, but without formal procedural 
investigations, they cannot alone be considered conclusive evidence. 
• 
Principle 2: Adopt a holistic approach to assessing racial and ethnic disparities by utilizing a 
variety of methods grounded in well-established scholarly literature. 
• 
Principle 3: Transparently outline the assumptions and limitations of each method, allowing 
policymakers and the public to make informed judgments based on each analysis.  
The report is structured to guide readers through a range of descriptive and statistical tests, each 
differing in its assumptions and scrutiny levels. This multi-test approach serves as a safeguard 
against potential errors, mitigating the possibility of (1) false positives- where a disparity is detected 
and where none exists (Type I error) and (2) false negatives- where a real disparity goes undetected 
(Type II error). For the analysis, demographic groups were categorized into four overlapping 
classifications to ensure a sufficiently large sample size for statistical validity. Although much of the 
focus is on stops involving Black and Hispanic individuals, the analysis also considers aggregated 
groups of all non-White individuals. In individual tests, we annotate results that were found to be 
statistically significant at a confidence level exceeding 90 percent. However, we consider highly 
significant results to be those exceeding a 95 percent confidence level.   
The analysis begins with the Solar Visibility analysis, a method developed by Grogger and Ridgeway 
(2006) to detect racial and ethnic disparities in stop data. This test focuses on stops occurring during 
the "inter-twilight window"—a fixed period each year where visibility fluctuates due to seasonal 
changes and daylight savings. It compares the ratio of non-White to White stops made in daylight 
versus darkness. The test assumes that if racial profiling occurs, troopers are more likely to act during 
daylight when race and ethnicity are easier to observe. By restricting the sample to this window and 
controlling for variables such as time of day and day of the week, any remaining differences in stop 
rates are attributed to potential disparate treatment. A higher likelihood of a non-White motorist 
having been stopped in daylight relative to darkness would indicate potential discriminatory 
behavior. 

3 
 
The second analytical tool used in the analysis is the synthetic control, where the number of non-
White traffic stops in a given department is evaluated against a benchmark constructed using stops 
made by all other departments in Connecticut. Since departments differ in terms of their enforcement 
activity (i.e., time of stops, reason for stops, etc.) and the underlying demographics of the population 
on the roadway, this analysis relies on the rich statistical literature on propensity scores. Here, a 
propensity score is a measure of how similar a stop made outside a given department is to a stop 
made by the department being analyzed. These measures of similarity are used to weight stops when 
constructing an individual benchmark for each department. For example, if the department being 
analyzed has a high non-White population and makes most of their stops on Friday nights at 7 PM for 
speeding violations, then stops made for speeding violations by departments with a similar 
residential population at this time and day will be given more weight when constructing the 
benchmark. This methodology ensures that there is an apples-to-apples comparison between the 
number of minorities stopped in a given town relative to their benchmark and allows for the 
interpretation of any remaining differences to be attributed to possible disparate treatment. 
Next, we examine disparities in traffic stop outcomes by analyzing the distribution of post-stop 
dispositions, conditional on race and the reason for the stop. Specifically, we test whether non-White 
individuals experience different outcomes compared to their White counterparts. In this section, we 
analyze differences in terms of arrests and precursors to an arrest, which we define as a discretionary 
search, vehicle exit, or field sobriety test. We condition a highly granular set of control variables to 
control for inherent differences that might necessitate a trooper engaging in discretionary post-stop 
enforcement. A higher rate of post-stop enforcement for non-White individuals relative to White 
individuals would indicate potential discriminatory behavior.  
 
Finally, the analysis explores post-stop outcomes using the hit-rate approach developed by Knowles, 
Persico, and Todd (2001). This approach assumes that individuals adjust their likelihood of carrying 
contraband in response to the probability of being searched, while police troopers make search 
decisions based on visible indicators of guilt. According to the model, a demographic group should 
only be searched more frequently than White non-Hispanic individuals if its members are more likely 
to carry contraband. However, the higher search rate should correspond exactly to a higher 
propensity to carry contraband. In the absence of racial bias, the success rate of searches (i.e., the hit 
rate) should be equal across all demographic groups. A lower hit rate for non-White individuals 
relative to White individuals would indicate potential discriminatory behavior. 
 
In summary, the analysis aims to identify statistically significant racial and ethnic disparities in traffic 
stop data. A variety of tests are applied to offer a comprehensive approach, incorporating lessons 
from both academic research and policy applications. Detailed explanations of each test's 
assumptions and mechanisms are provided to ensure policymakers and the public can assess the 
data and draw informed conclusions. Finally, we emphasize that these statistical tests can reveal 
racial and ethnic disparities indicative of racial/ethnic bias and potential discrimination, but without 
an additional procedural investigation, they cannot alone be considered conclusive evidence.

4 
 
II: CHARACTERISTICS OF TRAFFIC STOP DATA  
This section examines general patterns of traffic enforcement activities in Connecticut for the study 
period of January 1, 2023, to December 31, 2024. Statewide and agency activity information can be 
used to identify variations in traffic stop patterns to help law enforcement and local communities 
understand more about traffic enforcement. Although some comparisons can be made between 
similar communities, we caution against comparing agencies’ data in this report section. Please note 
that the tables in this report present information from only a few departments. Complete tables for 
all agencies are included in the technical appendix.   
In Connecticut, 357,852 traffic stops were conducted in calendar year 2023, and 413,137 traffic stops 
were conducted in calendar year 2024. Traffic enforcement increased by 14% in 2023 compared to 
2022 and 15% in 2024 compared to 2023. In both years, the month with the largest number of traffic 
stops occurred in April. In 2023, 73% of the total stops were conducted by the 95 municipal police 
departments, while the Connecticut State Police conducted 27%. In 2024, 77% of the total stops were 
conducted by the 95 municipal police departments, while the Connecticut State Police conducted 
23%. Figure 2.1 shows the aggregate number of monthly traffic stops and each demographic 
category. 
Figure 2.1: Aggregate Traffic Stops by Month of the Year 
 
The number of reported traffic stops declined steadily from 2014 to 2018, with an overall decrease 
of 16%. A slight uptick occurred in 2019, followed by a sharp drop in 2020—54% lower than 2019 
and 61% lower than 2014. Stops rebounded moderately in 2021 (up 17%) and again in 2022 (up 
15%), though levels remained below pre-pandemic figures. Growth continued through 2023 and 
2024 with increases of 14% and 15% respectively, compared to previous years. Figure 2.2 displays 
the total number of traffic stops by year since the start of the project. 
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
Jan-23
Feb-23
Mar-23
Apr-23
May-23
Jun-23
Jul-23
Aug-23
Sep-23
Oct-23
Nov-23
Dec-23
Jan-24
Feb-24
Mar-24
Apr-24
May-24
Jun-24
Jul-24
Aug-24
Sep-24
Oct-24
Nov-24
Dec-24
Black
Hispanic
All Other Stops

5 
 
Figure 2. 2: Connecticut, number of traffic stops, 2014-2024 
 
Figure 2.3 shows the aggregate number of traffic stops by month of the year between 2019 and 2024. 
Traffic stops have increased each month of each year since April 2020, but only one month 
(December 2024) has matched or exceeded pre-pandemic stop totals. Although the monthly trends 
remain similar, April and August contributed the largest number of traffic stops, and December had 
the lowest number of traffic stops in the calendar year.   
Figure 2. 3: Aggregate Traffic Stops by Month of the Year 2019 to 2024 
 
Figure 2.4 shows traffic stops by time of day over the analysis period. Stop volumes vary considerably 
by hour, with distinct peaks and lulls. In 2023, the highest hourly volume occurred between 5:00 and 
6:00 PM, accounting for 7.1% of all stops—consistent with peak commuting patterns in Connecticut. 
By contrast, the lowest volume occurred between 4:00 and 5:00 AM and remained low through the 
morning commute, likely reflecting an emphasis on maintaining smooth traffic flow and a reduced 
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
0
10000
20000
30000
40000
50000
60000
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
2020
2021
2022
2023
2024
2019

6 
 
likelihood of discretionary stops during these hours. In 2024, enforcement patterns shifted: the 
highest volume of stops occurred between 9:00 and 11:00 AM, comprising 15% of all stops. Notably, 
2024 marked the first year since this study began in 2015 that morning enforcement exceeded 
evening enforcement. 
Figure 2.4: Aggregate Traffic Stops by Time of Day 
 
 
Figure 2.5 displays traffic stops by day of the week for the entire analysis period. This figure shows 
that the number of traffic stops increases throughout the week and peaks between Tuesday and 
Thursday. Traffic stops decline substantially on the weekends, with the smallest number occurring 
on Sundays.  
Figure 2. 5: Traffic Stops by Day of Week 
 
0
5000
10000
15000
20000
25000
30000
35000
12-1:00 AM
1-2:00 AM
2-3:00 AM
3-4:00 AM
4-5:00 AM
5-6:00 AM
6-7:00 AM
7-8:00 AM
8-9:00 AM
9-10:00 AM
10-11:00 AM
11-12:00 PM
12-1:00 PM
1-2:00 PM
2-3:00 PM
3-4:00 PM
4-5:00 PM
5-6:00 PM
6-7:00 PM
7-8:00 PM
8-9:00 PM
9-10:00 PM
10-11:00 PM
11-12:00 AM
2023
2024
32,992
50,868
57,236
59,696
59,834
55,289
41,937
37,991
62,612
66,608
64,821
66,012
67,228
47,865
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
2023
2024

7 
 
Traffic stop enforcement levels and purposes vary widely across agencies, reflecting factors such as 
crash prevention in high-risk areas, responses to criminal activity, and citizen complaints. Agencies 
with dedicated traffic units generally record higher stop volumes. To standardize comparisons, traffic 
stops are measured per 1,000 residents. In 2023, the statewide rate was 127 stops per 1,000 
residents, rising to 146 in 2024. That year, 33 municipal departments exceeded the state rate, 
compared to 36 in 2024. In 2023, the highest stop rates were reported by Wilton, Orange, Waterford, 
Southington, and Ansonia, while the lowest were recorded in Shelton, Waterbury, Wolcott, 
Bridgeport, and Thomaston. In 2024, Waterford, East Windsor, Westport, Madison, and Wilton had 
the highest rates, while Shelton, Wolcott, Thomaston, Bridgeport, and Stratford had the lowest. Full 
results by department are provided in Tables A.1 (2023) and A.2 (2024) in Appendix A. 
Table 2.1 summarizes demographic characteristics of drivers stopped in Connecticut during the 
study period. Across both years, nearly two-thirds (63%) of stopped drivers were male, and the vast 
majority (88%) were Connecticut residents. About one-third of drivers were under age 30, while one-
quarter were over age 50. 
By race and ethnicity, White Non-Hispanic drivers accounted for the majority of stops, though their 
share declined slightly from 58.3% in 2023 to 57.7% in 2024. Black Non-Hispanic drivers made up 
18.7% of stops in 2023 and 18.2% in 2024. Hispanic drivers increased from 19.1% in 2023 to 20.4% 
in 2024, while drivers classified as Other (Asian/Pacific Islander and American Indian/Alaskan 
Native Non-Hispanic) remained stable at about 4% across both years. 
Table 2. 1: Statewide Driver Characteristics 
Race and Ethnicity 
2023 
2024 
Gender 
2023 
2024 
White 
58.3% 
57.7% 
Male 
63.2% 
63.6% 
Black 
18.7% 
18.2% 
Female 
36.8% 
36.4% 
Hispanic 
19.1% 
20.4% 
Age 
16 to 20 
8.5% 
8.2% 
Other 
3.8% 
3.7% 
21 to 30 
25.6% 
24.6% 
31 to 40 
22.8% 
22.8% 
Residency 
41 to 50 
16.8% 
17.2% 
CT Resident 
88.1% 
87.8% 
51 to 60 
14.0% 
13.9% 
Non-Resident 
11.9% 
12.2% 
Older than 61 
12.3% 
13.3% 
 
Police record the statutory reason for stopping a motor vehicle for every stop. Those statutes are 
then sorted into 15 categories, from speeding to registration and stop sign violations. For example, 
all statutory violations related to speed are categorized as speeding. Although speeding is the most 
often cited reason for stopping a motor vehicle statewide, the results vary by jurisdiction. Table 2.2 
presents the basis for traffic stops in 2023 and 2024. Speeding remained the most common reason 
for a stop, increasing slightly from 34.8% in 2023 to 35.7% in 2024. Stop sign violations (11.0% to 
11.1%) and traffic control signal violations (8.4% to 8.7%) were largely unchanged. 
Some categories saw modest shifts. Registration-related stops rose from 7.7% to 8.6%, while 
miscellaneous moving violations declined from 8.0% to 7.5%. Defective light violations also 
decreased, from 6.2% in 2023 to 5.5% in 2024. Seatbelt stops saw the largest drop, falling from 1.9% 
to 1.2%. Cell phone stops (6.3% to 6.1%), STC violations (3.0% to 2.8%), and display of plates (2.6% 

8 
 
to 2.3%) all declined slightly. The share of stops categorized as “All Other” increased marginally, from 
10.1% in 2023 to 10.5% in 2024. 
Table 2. 2: Basis for Stop  
Basis for Stop 
2023 
2024 
Speeding 
34.8% 
35.7% 
Defective Lights 
6.2% 
5.5% 
Misc. Moving Violation 
8.0% 
7.5% 
Stop Sign 
11.0% 
11.1% 
Registration 
7.7% 
8.6% 
Traffic Control Signal 
8.4% 
8.7% 
Cell Phone 
6.3% 
6.1% 
STC Violation 
3.0% 
2.8% 
Display of Plates 
2.6% 
2.3% 
Seatbelt 
1.9% 
1.2% 
All Other 
10.1% 
10.5% 
 
In addition to the difference in the volume of traffic stops across communities, agencies stopped 
drivers for various reasons. In 2023 and 2024, clear patterns emerge in the types of traffic violations 
most frequently cited by each agency statewide. Speed-related offenses dominate enforcement 
activity, accounting for approximately one-third of all citations issued by each department on 
average. Following speeding, violations, stop signs and traffic control signals violations consistently 
rank as the next most common categories for each agency. Moderate levels of enforcement are 
observed for moving violations, defective lights, and registration issues (typically between 7% and 
8% of enforcement by departments).  
At the departmental level, distinct outliers highlight the impact of local priorities. In both years, 
Weston consistently emerges as one of the most speed-focused departments, with speed-related 
violations exceeding 75% in 2023 and 82% in 2024.  Cell phone enforcement varies by department: 
Plymouth and Meriden reported higher levels of cell phone violations in 2023, while Hamden and 
Danbury reported the highest levels in 2024. These local differences underscore the importance of 
considering community-specific factors, such as traffic patterns, roadway design, and safety 
priorities, when evaluating enforcement outcomes across the state. 
Some Connecticut residents have expressed concern about the stops made for violations perceived 
as more discretionary, potentially making the driver more susceptible to possible police bias. Those 
stops are typically referred to as pretext stops and might include stops for defective lights, excessive 
window tint, or a display of plate violation, each of which, though a possible violation of state law, 
leaves the police officer with considerable discretion with respect to actually making the stop. A 
statewide combined average for stopping drivers for these violations is 11.5% in 2023 and 10.4% in 
2024.  While equipment-related violations make up a relatively small portion of overall traffic stops 
statewide, there is significant variation across departments. For most, equipment issues account for 
roughly one in ten citations; however, some departments place a much greater emphasis on this 
category. 
In 2023, the Mohegan Tribal Police stood out, with nearly 39% of all citations tied to equipment 
violations, more than three times the statewide average. Central Connecticut State University also 
shows a strong focus, with almost 29% of its citations in this category. Other departments with 

9 
 
elevated levels include Torrington (25.6%), Rocky Hill (24.8%), and Plainfield (24.8%), each 
recording equipment-related stops at roughly double the average. The pattern persists in 2024, 
where Central Connecticut State University once again reports a high concentration of equipment 
violations, this time at 30.8% of citations, reinforcing a consistent enforcement emphasis across both 
years. Additional departments with above-average enforcement include Plymouth (27.3%), 
Willimantic (25.7%), Ledyard (24.4%), and Plainville (22.6%), all of which issued equipment-related 
citations at more than twice the statewide average. All department results are contained in Tables 
A.3 (2023) and A.4 (2024) of Appendix A. 
In communities with a higher proportion of stops due to these violations, it is recommended that the 
departments be proactive in discussing the reasons for these stops with community members and 
examine for themselves whether or not such stops produce disparate enforcement patterns. 
Police generally stop a consistent percentage of drivers each month across most violation categories. 
For instance, administrative offenses account for roughly 3% of all stops each month, with little 
variation. However, four categories show more pronounced seasonal or monthly fluctuations. Cell 
phone violations typically represent about 6% of monthly stops, but enforcement spikes sharply in 
April (nearly 25%) and drops to just under 7% in October. These peaks likely correspond with 
federally funded distracted driving enforcement campaigns, which provide additional resources for 
targeted initiatives. Notably, the April surge in cell phone enforcement coincides with a decline in 
speeding stops, suggesting that such campaigns shift the focus of enforcement activity rather than 
increasing the total number of stops. 
Speeding violations peak in the summer months of July and August, while seatbelt enforcement 
follows a pattern similar to cell phone violations, with limited activity throughout most of the year 
but a significant increase in May, consistent with national seatbelt campaigns. In contrast, defective 
lighting violations demonstrate a seasonal trend tied to environmental conditions. Enforcement rises 
during the winter months, peaking between November and February, which corresponds with 
shorter daylight hours and a greater likelihood of lighting-related issues being detected. 
Figure 2.6 illustrates these patterns by showing the monthly distribution of stops for cell phone, 
speeding, seatbelt, and defective lighting violations. 
Figure 2. 6: Percent of Stops by Month for Select Violations 
 
 
 
 
 
 
 
 
 
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
45.0%
Jan.
Feb.
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Speeding
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
Jan.
Feb.
Mar.
Apr.
May
Jun.
Jul.
Aug.
Sept.
Oct.
Nov.
Dec.
Cell Phone

10 
 
 
 
 
 
 
 
 
 
 
The distribution of traffic stop outcomes remained largely consistent between 2023 and 2024, 
though there were modest shifts in enforcement emphasis. 
In 2023, the majority of stops resulted in warnings rather than infractions, summons, or arrest. 
Verbal warnings were the most common outcome, accounting for 38.1% of all stops, followed by 
written warnings (24.3%). Infractions represented nearly one-third of all stops (30.6%), while more 
serious outcomes such as misdemeanor summons (4.2%) and arrest (1.0%) were relatively rare. A 
small share of stops resulted in no disposition (1.7%) or vehicle searches (1.6%). 
In 2024, the general pattern persisted, but there were some notable shifts. Infractions increased to 
33.1%. At the same time, verbal warnings declined to 35.4%, though they still remained the single 
most frequent disposition. Written warnings increased slightly to 25.6%, while misdemeanor 
summons decreased to 3.1%. Arrests, no dispositions, and vehicle searches all remained relatively 
stable, with only minor changes from 2023. The stability in arrests and searches indicates no 
significant change in the most serious enforcement outcomes. Table 2.3 presents the outcomes of 
traffic stops in 2023 and 2024. 
Table 2. 3: Outcome of Traffic Stop  
Outcome of Stop 
2023 
2024 
Uniform Arrest Report 
1.0% 
0.9% 
Misdemeanor Summons 
4.2% 
3.1% 
Infraction Ticket 
30.6% 
33.1% 
Written Warning 
24.3% 
25.6% 
Verbal Warning 
38.1% 
35.4% 
No Disposition 
1.7% 
1.8% 
Vehicles Searched 
1.6% 
1.7% 
 
Some argue that it is difficult for police officers to identify the defining characteristics of a driver 
before initiating a stop. Yet, just as there are variations across departments in the reasons for traffic 
stops, there are also differences in the outcomes of those stops. These differences highlight the 
discretion local police departments have in enforcing state traffic laws. For example, some 
communities may prioritize issuing infraction tickets as the most effective way to promote traffic 
safety, while others may rely more heavily on warnings. This analysis provides police departments 
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
Jan.
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Seatbelt
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
9.0%
Jan.
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Defective Lights

11 
 
and local communities with a framework to understand how their enforcement activity compares to 
that of other jurisdictions. 
At the department level, outcome patterns shifted more noticeably than at the statewide level, 
underscoring the localized nature of enforcement practices. For instance, Fairfield, Waterbury, and 
CSP Troop H issued substantially more infractions in 2024, while Fairfield simultaneously reduced 
its use of verbal warnings. In contrast, Groton Long Point, East Lyme, and Easton relied far more on 
written warnings, with Groton Long Point increasing by more than 40 percentage points. Other 
departments moved in the opposite direction: Eastern Connecticut State University and Putnam 
issued fewer written warnings but leaned more heavily on verbal warnings. Wolcott exhibited 
significant fluctuations across categories, with increases in arrests and verbal warnings, but 
decreases in infractions and written warnings. Meanwhile, Waterbury, Shelton, and Norwich 
recorded sharp declines in the use of misdemeanor summonses. 
Although arrests remain rare in Connecticut, averaging 0.9% of stops in 2023 and 0.8% in 2024, some 
departments consistently report arrest rates several times higher than the statewide average. In 
2023, the highest rates were found in Willimantic (4.5%), CSP Troop H (4.5%), and New London 
(4.3%), with additional elevated levels in Bridgeport, Winsted, Yale University, and Groton City. By 
2024, arrest rates declined slightly, but departments such as Groton City, CSP Troop D, Bridgeport, 
and Plainfield remained well above average, along with Willimantic, CSP Troop H, New London, West 
Haven, and CSP Troop G. Full department-level outcomes are provided in Tables A.5 (2023) and A.6 
(2024) of Appendix A. 
Vehicle searches are also uncommon, occurring in only 1.6% of stops in 2023 and 1.7% in 2024. 
Nevertheless, significant variation exists across jurisdictions. When searches are conducted, officers 
must record their legal justification and whether contraband was recovered, providing an additional 
layer of accountability. A small number of departments consistently report search rates far above the 
statewide average. In both years, Waterbury, West Haven, and Bridgeport led the state, with search 
rates of 17–18% in Waterbury, around 11% in West Haven, and 10% in Bridgeport. Smaller 
jurisdictions, including Willimantic and Norwich, also exceeded the statewide average, though their 
rates moderated slightly in 2024. 
In addition to these consistently high-search departments, some agencies recorded substantial 
changes between the two years. Notable increases occurred in Wolcott (1.0% to 4.0%), New Canaan 
(0% to 2.9%), Monroe (1.5% to 4.0%), University of Connecticut (2.5% to 4.9%), and New London 
(3.3% to 5.1%). Conversely, the largest declines were observed in Yale University (3.6% to 0.4%), 
New Britain (6.0% to 2.9%), Willimantic (9.1% to 6.1%), Winsted (2.6% to 0.3%), and Middletown 
(3.4% to 1.1%). 
Overall, while the statewide averages for both arrests and searches remain low and stable, the 
department-level data reveal substantial variation in enforcement practices. Waterbury, West Haven, 
and Bridgeport consistently stand out as high-search jurisdictions, while other departments 
experienced notable shifts in their approach between 2023 and 2024. These differences reinforce the 
importance of examining enforcement not only at the statewide level but also within individual 
communities. All department-level results are provided in Tables A.7 (2023) and A.8 (2024) of 
Appendix A. 

12 
 
III: ANALYSIS OF TRAFFIC STOPS, SOLAR VISIBILITY 
Assessing racial and ethnic disparities in police decisions to stop motorists presents a significant 
challenge, primarily due to the absence of an appropriate counterfactual. In simple terms, reliable 
data on the demographic composition of motorists is unavailable, leaving no benchmark for 
evaluating police traffic stops. To address this challenge, we implement a solar visibility analysis, 
following the methodology established by Grogger and Ridgeway (2006). The authors demonstrate 
that, under specific conditions, a change in the odds of a stopped motorist being a minority from 
daylight to darkness is equivalent to a change in the odds that a minority motorist is stopped. 
Assuming the only variable changing between daylight and darkness is the police's ability to detect 
race before making a traffic stop, an increased likelihood of stopping a minority motorist during 
daylight indicates disparate treatment and possible discrimination (see also Ridgeway 2009; Horrace 
and Rohlin 2019; Kalinowski et al. 2018, 2020, 2022).2 Ross et al. (2020) note that when they wrote 
their study, 18 cities and four states had already adopted solar visibility analysis to evaluate traffic 
stops. 3 They describe this approach as being widely considered a “best practice” among practitioners 
and policymakers, a view that has gained even more traction since the study