Roundabout Crash Analysis in Low versus High Income Areas in Southeast Michigan (2019–2023)

Haashir Ahmed; Aarit Das; Haashir Ahmed; Aarit Das

doi:10.3998/umurj.9833

Introduction

Transportation inequality is a major contributor to urban inequities in the United States. The inequality across transportation infrastructure continues to reflect the long-standing history of institutional racism, where socioeconomically disadvantaged areas^¹

“Socioeconomically disadvantaged areas” will be the terminology used throughout this analysis to collectively describe communities with lower income, higher proportions of people of color, or both.

(typically communities of color) have fewer resources to implement and utilize newer, safer traffic systems.

Roundabouts offer an ideal example: they are one of the most ubiquitous ways of improving traffic safety in modern roadway systems, helping improve traffic safety and efficiency through slowing down vehicles, increasing traffic volume, and reducing conflict points, leading to fewer crashes (39). Additionally, roundabouts provide significant economic benefits due to their lack of traffic signals or signs, as well as environmental benefits by improving traffic flow and reducing the need for vehicles to idle (35; 3). However, for all their benefits, roundabouts are less prevalent in socioeconomically disadvantaged areas, which tend to have more outdated traffic signal systems instead (4).

This inequality stems from the effects of historically discriminatory planning strategies, such as Redlining and urban renewal projects, which have hampered disadvantaged communities for decades. To this day, due to their proximity to large highways and roads, Black^²

“Black” is used throughout this analysis to collectively describe individuals who identify as Black or African-American.

and Latino communities are exposed to 56% and 63% more pollution, respectively, and urban areas with no White residents have a 7% chance of having well-maintained roads, compared to the 22% of roads in nearly all-White neighborhoods (46; 24). Additionally, disadvantaged and minority communities also tend to experience poorer road quality and fewer reliable transportation options, while facing longer commute times, all of which exacerbate economic challenges in disadvantaged communities (48; 25; 34).

This raises an important question: if newer transportation methods were implemented equitably across communities, regardless of socioeconomic factors, would disadvantaged communities, where crash and injury rates are already higher, experience significant benefits in crash safety (19; 49; 50)? At first glance, the benefits of such interventions seem straightforward; however, existing studies have shown that traffic interventions vary in effectiveness depending on environmental and traffic factors (2). They tend to only benefit significantly if implemented on a broader scale with complementary changes to surrounding traffic systems (18). This insight encourages an in-depth examination of whether modern traffic control systems can improve traffic safety across communities regardless of socioeconomic levels.

Research Question

This analysis will focus on how modern traffic control designs, specifically roundabouts, affect crash rates in Southeast Michigan between 2019 and 2023, based on Southeast Michigan Council of Governments (SEMCOG) and OpenStreetMap data, while accounting for income variations using the Integrated Public Use Microdata Series National Historical Geographic Information System (IPUMS NHGIS) Census Data. While prior research has examined the benefits of modern traffic control designs, few have focused on these benefits across varying socioeconomic levels. By linking crash data with socioeconomic information, this study aims to fill that gap, determining whether modern traffic infrastructure can truly equitably improve traffic safety.

The analysis will focus on seven of Michigan’s Southeast counties— Livingston, Macomb, Monroe, Oakland, St. Clair, Washtenaw, and Wayne. These include three of the largest counties in Michigan by population size (Wayne, Oakland, and Macomb). These seven counties also include some of Michigan’s wealthiest counties— Washtenaw, Livingston, and Oakland, ranking 2nd, 3rd, and 4th, respectively, in median income in Michigan in 2023— while including Wayne, Michigan’s second poorest county by percentage of residents under the poverty line (44; 20). These factors make these seven counties ideal for incorporating a wide range of socioeconomic levels.

These seven counties also uniquely represent Michigan as one of the earliest adopters of roundabouts in the United States, with the first one implemented in 1999 in Oakland (9; 45). Since then, Michigan has experienced rapid implementation of roundabouts, with their implementation reflecting the proactive approach to traffic safety of wealthier counties, such as Oakland, which has the most roundabouts of any county in Michigan (14). These variations in implementation across counties provided a natural context for studying traffic safety and roundabouts in Michigan.

Hypotheses

H1:
Roundabouts reduce crashes of all types equally, regardless of their location in higher or lower-income areas.
H2:
Roundabouts disproportionately reduce crashes of all types in higher- income neighborhoods due to improved transportation systems.
H3:
Roundabouts disproportionately reduce crashes of all types in lower- income neighborhoods due to their self-enforcing design.
H4:
Roundabouts substantially reduce crash types that disproportionately affect lower-income communities (17; 23).

Literature Review

1. Historical Transportation Inequality and Redlining:

Transportation inequality has been the subject of extensive research, with numerous studies demonstrating the long-lasting effects of historical transportation inequality on current systems. One of the major contributors to this inequality is redlining, a discriminatory policy originating from government-backed maps that marked minority neighborhoods as high-risk for investment. This allowed banks to deny loans or insurance for residents, typically those of color, from these neighborhoods (32). To this day, tracts that were graded “D” or hazardous in the 1930s, under Redlining policies, have an increased rate of present-day pedestrian fatalities compared with tracts graded “A” (47). Additionally, tracts marked as “D” have more than double the pedestrian fatality rate (2.6 per 100,000 people) of tracts marked “A.” Redlining institutionalized racism almost a century ago, yet the effects of its policies are felt by millions, some 50 years after it was officially outlawed.

2. Racial and Socioeconomic Disparities in Traffic Safety:

Due to the historical inequality within the traffic system in the United States, traffic safety is highly dependent on racial and socioeconomic factors. For example, in Saint Louis, Missouri, street blocks with a majority Black population tend to have a higher occurrence of uneven sidewalks (21). Meanwhile, Black residents tend to rely on public transportation five times more to access vital services and opportunities due to a lack of personal vehicles. On top of this, there are 60% more pedestrian fatalities for Black than White pedestrians (13). Transportation access (or lack thereof) for healthcare is also a primary reason why Black Americans have been disproportionately affected^³

Out of every 100,000 Black Americans, 97.9 have died from COVID-19, a fatality rate more than double that of the fatality rate for White Americans, which is only 46.4 per 10,0,000 people (5).

by COVID-19.

These disparities draw directly from the creation of the modern United States interstate system, which saw the targeted demolition of entire neighborhoods of color for highway expansions, displacing hundreds of thousands of Black Americans (26; 30). The resulting highways also acted as physical barriers, enabling the segregation of communities of color and isolating them from economic hubs such as downtowns (1).

Consequently, these structural inequities continue to shape current traffic safety. For instance, a higher income level among vehicle owners is significantly correlated to greater access to vehicle safety features (28). It may, therefore, be no surprise that disadvantaged communities experience higher injury counts, even after accounting for traffic exposure (11). At the same time, disadvantaged communities often have poorer traffic conditions with more speeding, denser traffic, and worse infrastructure (8). This, in turn, results in more crashes in urban arterial roads and socioeconomically disadvantaged communities. Nonetheless, research has purported that these discrepancies are solvable if multifaceted interventions are implemented on a system-wide scale (8.

3. Modern Traffic Safety Frameworks and Equity Gaps:

Traffic safety has traditionally been framed under the assumption that crashes are largely unavoidable and are primarily the fault of individual road users, particularly pedestrians or cyclists (16). This assumption has allowed traffic safety to be defined around individual accountability rather than system improvement (51). It is this perspective that most modern-day transportation safety programs, such as Vision Zero and the Safe System Approach (SSA), are of relevance. Vision Zero refers to a traffic safety framework, centered on shifting the perspective that traffic fatalities/injuries are inevitable to recognizing that they are preventable through system redesign (16). Meanwhile, while SSA is closely aligned with Vision Zero, it aims to decrease crash fatalities/injuries through prioritizing system redundancy and fail-safe mechanisms in vehicles and roadway designs (38). However, while both Vision Zero and SSA are steps in the right direction, they typically do not systematically account for the historical impacts of structurally racist policies and inequities related to transportation safety (42; 43).

4. Roundabouts and Crash Reduction:

This analysis focuses extensively on roundabouts and their corresponding effects on crash rates in lower and higher-income communities. There has been ample literature on the effects of roundabouts on crash rates, pointing towards significant decreases in crash and injury rates at roundabouts. For instance, a study in California examining 24 intersections that were converted to roundabouts found a 76% decrease in injury crashes (29). Meanwhile, based on the results of a 2004 study, it is estimated that converting 10% of intersections into roundabouts in the United States could have prevented approximately 50,000 crashes in 2023 (36; 10). These and many more sources point towards the incredible effectiveness of roundabouts (12; 31; 40). However, a notable gap is present in the literature regarding the varying effects of roundabouts in high and low-socioeconomic areas.

Methodology

This analysis compares roundabouts to similar intersections, which are defined based on key structural and traffic characteristics. The study is divided into two parts: P1, which identifies roundabouts, and P2, which identifies comparable non-roundabout intersections.

P1 focuses on roundabouts that meet criteria (ω): constructed before 2019^⁴

As explained later in the Methodology, the study period is between 2019 and 2023. Thus, roundabouts built before 2019 were excluded to ensure the constant crash rate comparison for all roundabouts during the five-year study period.

and located in one of the seven counties of interest, Livingston, Macomb, Monroe, Oakland, St. Clair, Washtenaw, or Wayne. This is accomplished by identifying all roundabouts that meet criteria, using the Kittelson & Associates Roundabouts Database, which includes various data such as location, latitude, longitude, type, number of approaches, functional class, and year completed for each roundabout (37).

Each of the roundabouts was subsequently verified using Google Maps to ensure accuracy. This process resulted in a final sample size of 101 roundabouts that met the criteria. While 101 roundabouts met the criteria, only a subset were correctly identified during the data analysis (around 80%). These roundabouts are distributed across several counties as summarized in Table 1.

Table 1:

Number of Roundabouts in each County^⁵

All information, including coordinates and year built, can be found in Appendix A.

County	Washtenaw	Wayne	Oakland	Monroe	Macomb	Livingston	St. Clair
Number of Roundabouts	29	4	38	2	15	12	1

P1 is built upon SEMCOG crash data, gathered from Michigan State Police, and socioeconomic data from the National Historical Geographic Information System (NHGIS) 2023 American Community Survey (41, 27). The SEMCOG crash data denotes all crashes in Michigan’s southeast counties from 2015 to 2024. However, since the NHGIS dataset only covers the period from 2019 to 2023, the SEMCOG crash data was filtered to only include crashes between 2019 and 2023.

The SEMCOG data^⁶

The key demographic data mainly included the annual household income of the census tract in which the crash occurred in.

includes the coordinates of each crash, allowing for each crash to be assigned key demographic data from the census tract in which it has occurred. Next, using the latitude and longitude from the center^⁷

‌The coordinates for the center of the roundabout are given in the Kittelson & Associates Roundabouts Database, and listed for each roundabout used in this study in Apendix A.

of a roundabout, the crashes occurring within a 75-meter radius of the roundabout’s center are assigned to the roundabout. Crashes associated with each roundabout are aggregated across multiple variables^⁸

Variables are crash severity levels and contributing-factor for different crashes occurring at a given roundabout.

using individual crash records from the SEMCOG dataset. Crash rates are then calculated using a per-Million Entering Vehicles (MEV) metric. MEV is calculated through:

MEV = (AADTR (75th Percentile) * days (365) * study period (5)) / 1,000,000

The MEV is calculated using the 75th percentile^⁹

The 75th percentile for the AADTR is used instead of the mean to emphasize how the majority of crashes occur on the higher-volume streets that constitute the majority of crashes on roads leading to a roundabout.

of the AADTR^¹⁰

The average volume of vehicles passing through a specific road section or intersection per day, calculated over a year. It serves as the base number for determining how ‘busy’ an intersection is.

(Annual Average Daily Traffic) times the number of days for the study period of 5 years. Next, the crash rates for each variable are determined by dividing the crash rates for each variable by the MEV. This process is repeated for all crash variables and roundabouts included in the analysis.

The similar intersections were identified using roadway centerline data from OpenStreetMap (OSM) data (7). Using the OSM data, the start and end points of each road segment were identified. Locations where two or more road segment endpoints met were classified as intersections. Comparison intersections were limited to only those whose mean AADTR fell within the mean AADTR^¹¹

Mean AADTR was calculated as the mean of the AADTR values of all crashes occurring at each intersection.

range of the roundabouts (928.571 to 51,871). Additionally, similar to the roundabouts, each intersection included crashes within 75 meters of its center. Crashes were from the same SEMCOG data, combined with census-tract socioeconomic data, used for roundabouts. Crash rates for these intersections were calculated using the same (MEV) methodology applied to the roundabout sites, ensuring consistency.

The issue with this approach is double-dipping. This can happen when an intersection or roundabout is located within 150 meters of another intersection or roundabout, causing the 75-meter buffers to overlap and allocating the same crash to multiple locations.To avoid this, intersections or roundabouts within the same buffer area were dropped from the final analysis, retaining only the roundabout or intersection with the highest mean AADTR within each overlapping buffer area.

Intersections and roundabouts are then split into high or low-income communities. This is accomplished by determining if the median household income of the roundabout or intersection is below or above a certain threshold. In this analysis, $71,149 was used as the specific threshold to classify high versus low-income neighborhoods, as this is the median household income in Michigan (22).

Data Analysis

The analysis of the figures (Appendix B) shows us a complex relationship between intersection designs, socioeconomic status, and driver behavior. Collectively, these figures convey that while roundabouts are a powerful tool for improving safety, their effectiveness is influenced by the socioeconomic environment in which they are placed. In the figures, we see two different types of narratives (operational challenges and behavioral choices). Operational challenges are seen in Figures 1–6, with low-income communities experiencing a disproportionate number of No Factor and Hit-and-Run crashes at roundabouts. The No Factor^¹²

A crash where none of the specific contributing factors in the SEMOC data-set are marked as present. These are typically routine accidents (like minor rear-ends in congestion) caused by traffic flow issues rather than reckless behavior like speeding or impairment.

rate in low-income communities (4.54 MEV) is roughly double that of the high-income communities (2.49 MEV). This means that factors such as higher congestion, older vehicle maintenance, or less familiarity with roundabout navigation create a gap in operational safety. This is also demonstrated in the Hit-and-Run data, where roundabouts in low-income communities have a rate of 0.694 MEV compared to 0.336 MEV in wealthier areas, showcasing a critical disparity in accountability and system-wide safety.

For behavioral errors as seen in Figures 2, 3, 4, and 5, high-income communities show a higher prevalence of crashes linked to reckless behavior at roundabouts. Most notably, the speeding crash rate in roundabouts in high-income communities is approximately 123% higher than in the low-income communities, 0.209 MEV and 0.094 MEV, respectively, which is a sharp reversal of the trends seen at standard intersections. This suggests that drivers in wealthier areas may navigate modern infrastructure more aggressively. Similarly, roundabouts in high-income communities show higher rates for alcohol and drug-related incidents. This serves as an indicator of how wealth does not insulate a community from impairment-driven safety issues.

In all, when viewed in combination, these figures demonstrate that roundabouts significantly alter standard traffic safety. At traditional intersections, low-income communities generally perform worse across almost all categories. However, at roundabouts, the data shows that low-income communities actually perform better than high-income communities in categories like speeding and alcohol involvement. This suggests that the self-enforcing design of roundabouts (which is meant to force drivers to slow down their vehicles) may help benefit lower-income communities by reducing high-speed and high-severity crashes.

Results

The data from the roundabouts across Southeast Michigan indicates that roundabouts generally experience higher normalized crash rates than traditional intersections. However, the nature of these crashes varies significantly based on the neighborhood income. For instance, low-income communities tend to experience more No Factor crashes at roughly a rate of 4.54 MEV, which is more than double the rate as seen in roundabouts in high-income communities (2.49 MEV). This is also far higher than traditional similar-intersections (high-income: 0.26 MEV and low-income: 0.33 MEV). Conversely, roundabouts in high-income communities show a higher prevalence of behavioral errors; they recorded a speeding crash rate of 0.21 MEV, which is roughly 123% higher than the 0.09 MEV rate in roundabouts in low-income communities, and an alcohol-related crash rate that is nearly 20% higher (0.073 MEV versus 0.061 MEV). Distracted driving and drug-related incidents also followed this counterintuitive trend, with high-income communities showing slightly higher rates at roundabouts than low-income communities. However, the most profound socioeconomic disparity exists in Hit-and-Run incidents, where roundabouts in low-income communities suffer a rate of 0.69 MEV, roughly double the 0.34 MEV recorded in wealthier areas.

In summary, while roundabouts in low-income communities showed improvements in behavior-related crashes compared to roundabouts in high- income communities, they still faced a higher overall crash rate. This was demonstrated by the No Factor crash rate, which occurred at roughly double the rate in roundabouts in low-income communities than in roundabouts in high-income communities. This suggests that operational challenges, such as higher congestion and older vehicle maintenance, offset the behavioral safety improvements.”

Hypothesis Evaluation:

H1
(equal reduction in crashes across high and low income) was not supported, with crash rates showing significant differences per income level
H2
(greater reduction for crashes in high-income communities) was partially supported, with overall crash rates lower in higher-income areas. However, higher-income areas also saw an increase in crash rates for behavioral crashes.
H3
(greater reduction in crashes in low-income communities) was partially supported, with behavioral crashes occurring to a lesser extent in low- income communities, but in general, crash rates in low-income communities were significantly higher.
H4
(greater reduction in crashes disproportionately affecting low-income communities) was proven correct, with speeding and behavioral crashes occurring to a lesser extent in low-income communities.

Conclusion

This study highlights a complex pattern. In general, roundabouts in low-income communities experience a significantly higher overall crash rate (such as No Factor crashes). However, roundabouts in low-income communities also show improved rates of crashes related to behavioral factors (such as speeding and alcohol) than with roundabouts in high-income communities.

Limitations

The findings of this analysis have several limitations regarding the timeframe in which we gathered this data. The timeframe was only 5 years, and it is possible that with more data over a broader timeframe, the results could have been substantially different. Additionally, the crash data utilized a 5-year timeframe, which may be too small to capture long-term safety trends. This is a concern for roundabouts constructed just before and included in the study period (such as 2017 or 2021), as research indicates that road users require a transitional period to adapt to new traffic configurations (6). Furthermore, the study couldn’t account for specific intersections or roundabouts undergoing renovations during the study period, which may have affected crash frequencies. The study links crash locations to local socioeconomic data; however, crashes may involve non- residents from different areas who are more likely to experience accidents due to unfamiliarity with roundabout navigation. Finally, while a threshold of approximately $71,149 was used based on standard 4-person household income models, the NHGIS data represents the total income of all residents regardless of specific household.

For roundabouts and intersections where buffers overlapped, only the roundabout or intersection with the highest mean AADTR was retained. This may bias the analysis toward higher-traffic locations and exclude nearby lower-traffic sites. Finally, out of all the 101 roundabouts identified, only ~80% were successfully linked to crashes using the spatial method, which resulted in an analysis with only 78 of the 101 roundabouts identified. This occurred because of small offsets between the roundabout center coordinates and the crash point locations, or roundabouts without valid geometry.

Nonetheless, the biggest surprise is how, across the board, roundabouts had higher crash rates for both high and low-income communities compared to similar intersections for counties in southeastern Michigan. This directly contradicts existing literature showing that roundabouts reduce crashes; however, this matches previous analyses in Michigan showing a higher crash rate at roundabouts, but with less severe crashes (which may explain why No Factor crashes were so high) (15). This discrepancy may also stem from the adaptation period needed for drivers to adjust to roundabouts, with the majority of the crashes examined in this study occurring just a couple of years before the study period. Thus, future research should also examine crash severity rather than frequency alone, as roundabouts may reduce high-severity crashes while increasing minor incidents.

Acknowledgements

The use of AI to assist with programming and streamlining some of the intensive data cleaning programs. SEMCOG, OSM, and NHGIS for providing their datasets free of charge to those conducting analyses such as ours. Professor Attiya Shaw for allowing one of the coauthors of this paper to intern in her Infrastructure for All lab in 2025, providing the coauthor with much of the knowledge in civil engineering behind this paper.

References

1. 1956 – Federal-Aid Highway Act Displaces Black Americans - Florida Policy Timeline. (2025). Florida Policy Timeline. https://www.floridatimeline.org/timeline/1956-federal-aid-highway-act-displaces-black-americans/#:~:text=The%20highways%20also%20contributed%20to,underfunding%20in%20predominantly%20Black%20neighborhoods https://www.floridatimeline.org/timeline/1956-federal-aid-highway-act-displaces-black-americans/#:~:text=The%20highways%20also%20contributed%20to,underfunding%20in%20predominantly%20Black%20neighborhoods

2. Akbari, M., Heydari, S. T., Razzaghi, A., Vali, M., Tabrizi, R., & Lankarani, K. B. (2024). Effectiveness of interventions for preventing road traffic injuries: A systematic review in low-, middle- and high-income countries. PLoS ONE, 19. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0312428 https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0312428

3. Benefits of Roundabouts. (n.d.). FDOT. https://www.fdot.gov/agencyresources/roundabouts/benefits.shtm https://www.fdot.gov/agencyresources/roundabouts/benefits.shtm

4. Bhagat-Conway, M., Hernandez, A. C., Palm, M., & Combs, T. S. (n.d.). Roundabouts are for rich people: Who gets traffic control devices and what to do about it.

5. Christie, N., Jones, S., & O’Toole, S. E. (2025). Systemic inequalities in road safety outcomes across high income countries and lessons from intervention approaches. Journal of Transport & Health, 41, 102006–102006. https://doi.org/10.1016/j.jth.2025.102006 https://doi.org/10.1016/j.jth.2025.102006

6. Dong, S., Hussain, S. H., Chen, F., Zhou, J., Xu, F., & Khattak, A. (2023). Driver Adaptability When Traffic Side Is Switched from Left to Right and Vice Versa: A Driving Simulator Study with Chinese and Pakistani Drivers. Sustainability, 15(23), 16247–16247. https://doi.org/10.3390/su152316247 https://doi.org/10.3390/su152316247

7. Download OpenStreetMap for Michigan | Geofabrik Download Server. (2026). Geofabrik Download Server. https://download.geofabrik.de/north-america/us/michigan.html https://download.geofabrik.de/north-america/us/michigan.html

8. Dumbaugh, E., Li, Y., Saha, D., & Marshall, W. (2022). Why do lower-income areas experience worse road safety outcomes? Examining the role of the built environment in Orange County, Florida. Transportation Research Interdisciplinary Perspectives, 16, 100696.https://doi.org/10.1016/j.trip.2022.100696 https://doi.org/10.1016/j.trip.2022.100696

9. Ealy, J., Bramble, M., Clover, A., Kanitz, D., Mcdonnell, R., & Mcquiston, C. (n.d.). Project Information REPORT NAME: Evaluating the Performance and Safety Effectiveness of Roundabouts- MDOT Project Manager. Retrieved January 13, 2026, from https://www.michigan.gov/mdot/-/media/Project/Websites/MDOT/Programs/Research-Administration/Research-Spotlights/SPR-1725-Spotlight.pdf https://www.michigan.gov/mdot/-/media/Project/Websites/MDOT/Programs/Research-Administration/Research-Spotlights/SPR-1725-Spotlight.pdf

10. Eisenman, S., Josselyn, J., List, G., Persaud, B., Lyon, C., Robinson, B., Blogg, M., Waltman, E., & Troutbeck, R. (2004). OPERATIONAL AND SAFETY PERFORMANCE OF MODERN ROUNDABOUTS AND OTHER INTERSECTION TYPES. https://www.dot.ny.gov/divisions/engineering/technical-services/trans-r-and-d-repository/C-01-47.pdf https://www.dot.ny.gov/divisions/engineering/technical-services/trans-r-and-d-repository/C-01-47.pdf

11. Girasek, D. C., & Taylor, B. (2010). An Exploratory Study of the Relationship Between Socioeconomic Status and Motor Vehicle Safety Features. Traffic Injury Prevention, 11(2), 151–155. https://doi.org/10.1080/15389580903531598 https://doi.org/10.1080/15389580903531598

12. Gross, F., Lyon, C., Persaud, B., & Srinivasan, R. (2013). Safety effectiveness of converting signalized intersections to roundabouts. Accident Analysis & Prevention, 50, 234–241. https://doi.org/10.1016/j.aap.2012.04.012 https://doi.org/10.1016/j.aap.2012.04.012

13. Heller, M. (2021, April 22). The next step for antiracism is transportation - here’s why. World Economic Forum. https://www.weforum.org/stories/2021/04/transport-us-antiracism/https://www.weforum.org/stories/2021/04/transport-us-antiracism/

14. Hicks, J. P., & Levin, S. (2023, October 9). Michigan continues to add roundabouts. See which have the most crashes. Mlive. https://www.mlive.com/public-interest/2023/10/michigan-continues-to-add-roundabouts-see-which-have-the-most-crashes.html#:~:text=Of%20Michigan’s%2083%20counties%2C%20almost,Rapids%20and%20Kalamazoo%20have%20six https://www.mlive.com/public-interest/2023/10/michigan-continues-to-add-roundabouts-see-which-have-the-most-crashes.html#:~:text=Of%20Michigan’s%2083%20counties%2C%20almost,Rapids%20and%20Kalamazoo%20have%20six

15. Hunter, G. (2025). Roundabouts have most car accidents in Detroit area, but experts see upside. The Detroit News. https://www.detroitnews.com/story/news/local/michigan/2024/06/09/roundabouts-have-most-car-accidents-in-detroit-area-but-experts-see-upside/73713496007/https://www.detroitnews.com/story/news/local/michigan/2024/06/09/roundabouts-have-most-car-accidents-in-detroit-area-but-experts-see-upside/73713496007/

16. It is a system failure in decision making rather than a failure of individuals. (2026). Carbonindependent.org. https://www.carbonindependent.org/38.html#:~:text=The%20traditional%20approach%20to%20road%20safety%2C%20of,not%20behave%20as%20expected%20by%20road%20engineers.https://www.carbonindependent.org/38.html#:~:text=The%20traditional%20approach%20to%20road%20safety%2C%20of,not%20behave%20as%20expected%20by%20road%20engineers

17. Kazuyuki Neki, Mitra, S., Wambulwa, W. M., & Franklin, R. (2022). Profile of low and middle-income countries with increases versus decreases in road crash fatality population rates and necessity of motorcycle safety. Journal of Safety Research, 84, 129–137. https://doi.org/10.1016/j.jsr.2022.10.014 https://doi.org/10.1016/j.jsr.2022.10.014

18. Khan, M. N., & Das, S. (2024). Advancing traffic safety through the safe system approach: A systematic review. Accident Analysis and Prevention, 199. https://www.sciencedirect.com/science/article/abs/pii/S0001457524000630?via%3Dihub https://www.sciencedirect.com/science/article/abs/pii/S0001457524000630?via%3Dihub

19. Kyoon Yoo, D., & Professor, A. (n.d.). Relationship Between Socioeconomic Inequality and Traffic Crashes on Virginia Roads at the County and Census-Tract Levels. https://library.virginiadot.org/vtrc/supplements https://library.virginiadot.org/vtrc/supplements

20. Lock. (2026). Nih.gov. https://hdpulse.nimhd.nih.gov/data-portal/social/table?socialtopic=080&socialtopic_options=social_6&demo=00007&demo_options=poverty_3&race=00&race_options=race_7&sex=0&sex_options=sexboth_1&age=001&age_options=ageall_1&statefips=26&statefips_options=area_states https://hdpulse.nimhd.nih.gov/data-portal/social/table?socialtopic=080&socialtopic_options=social_6&demo=00007&demo_options=poverty_3&race=00&race_options=race_7&sex=0&sex_options=sexboth_1&age=001&age_options=ageall_1&statefips=26&statefips_options=area_states

21. Lowe, K. (2016). Environmental Justice and Pedestrianism: Sidewalk Continuity, Race, and Poverty in New Orleans, Louisiana. Transportation Research Record Journal of the Transportation Research Board, 2598(1), 119–123. https://doi.org/10.3141/2598-14 https://doi.org/10.3141/2598-14

22. Michigan | Data USA. (2023). Datausa.io. https://datausa.io/profile/geo/michigan https://datausa.io/profile/geo/michigan

23. Mitsova, D., Dumbaugh, E., & Saha, D. (2021). Crash Risk for Low-Income and Minority Populations: An Examination of At-risk Population Segments and Underlying Risk Factors Acknowledgment of Sponsorship. https://www.roadsafety.unc.edu/wp-content/uploads/2022/01/R31-Low-Income-Age-Gender-Cohorts-Final-Report.pdf https://www.roadsafety.unc.edu/wp-content/uploads/2022/01/R31-Low-Income-Age-Gender-Cohorts-Final-Report.pdf

24. National Highways: Analysis of Available Data Could Better Ensure Equitable Pavement Condition. (n.d.). https://www.gao.gov/products/gao-22-104578 https://www.gao.gov/products/gao-22-104578

25. Ney, J. (n.d.). Long commutes show structural inequality in cities. Https://Jeremybney.Medium.Com/Long-Commutes-Show-Structural-Inequality-in-Cities-1b55f460f26a Https://Jeremybney.Medium.Com/Long-Commutes-Show-Structural-Inequality-in-Cities-1b55f460f26a

26. O’Hare, L. M. (2024). Concrete Legacy: The Effects of the Interstate Highway System on Black Communities in the U.S. DigitalCommons@Macalester College. https://digitalcommons.macalester.edu/tapestries/vol13/iss1/6/https://digitalcommons.macalester.edu/tapestries/vol13/iss1/6/

27. Overview of NHGIS Datasets | IPUMS NHGIS. (2023). Nhgis.org. https://www.nhgis.org/overview-nhgis-datasets https://www.nhgis.org/overview-nhgis-datasets

28. Pirdavani, A., Daniels, S., Vlierden, K. van, Brijs, K., & Kochan, B. (2016). Socioeconomic and sociodemographic inequalities and their association with road traffic injuries. Journal of Transport & Health, 4, 152–161. https://doi.org/10.1016/j.jth.2016.12.001 https://doi.org/10.1016/j.jth.2016.12.001

29. Problem: Intersection crashes account for more than 45 percent of all crashes nationwide. (n.d.). https://highways.dot.gov/media/9206 https://highways.dot.gov/media/9206

30. Racism has shaped public transit, and it’s riddled with inequities | Kinder Institute for Urban Research. (2020). Kinder Institute for Urban Research | Rice University. https://kinder.rice.edu/urbanedge/racism-has-shaped-public-transit-and-its-riddled-inequities https://kinder.rice.edu/urbanedge/racism-has-shaped-public-transit-and-its-riddled-inequities

31. Raised Crosswalks on Entrance to the Roundabout—A Case Study on Effectiveness of Treatment on Pedestrian Safety and Convenience. (2026). Traffic Injury Prevention. https://doi.org/10.1080//15389588.2013.854885 https://doi.org/10.1080//15389588.2013.854885

32. redlining. (2022). LII/Legal Information Institute. https://www.law.cornell.edu/wex/redlining https://www.law.cornell.edu/wex/redlining

33. Reyes, M. V. (2020). The Disproportional Impact of COVID-19 on African Americans. Health and Human Rights, 22(2), 299. https://pmc.ncbi.nlm.nih.gov/articles/PMC7762908/https://pmc.ncbi.nlm.nih.gov/articles/PMC7762908/

34. Roberto, E. (2008). “Among house-Commuting to Opportunity: The Working Poor and Commuting in the United States.

35. Roundabouts | FHWA. (n.d.). https://highways.dot.gov/safety/intersection-safety/intersection-types/roundabouts https://highways.dot.gov/safety/intersection-safety/intersection-types/roundabouts

36. Roundabouts. (2024). IIHS-HLDI Crash Testing and Highway Safety. https://www.iihs.org/research-areas/roundabouts#:~:text=Studies%20of%20intersections%20in%20the,primarily%20on%20single%2Dlane%20roundabouts https://www.iihs.org/research-areas/roundabouts#:~:text=Studies%20of%20intersections%20in%20the,primarily%20on%20single%2Dlane%20roundabouts

37. Roundabouts Database Home. (2021). Kittelson.com. https://roundabouts.kittelson.com/https://roundabouts.kittelson.com/

38. Safe System Approach. (2026). Michigan.gov. https://www.michigan.gov/mdot/travel/safety/efforts/safe-system-approach https://www.michigan.gov/mdot/travel/safety/efforts/safe-system-approach

39. Safety/Other Roundabout Benefits | Nevada Department of Transportation. (n.d.). https://www.dot.nv.gov/safety/roadway-safety-improvements/roundabouts/safety-other-roundabout-benefits https://www.dot.nv.gov/safety/roadway-safety-improvements/roundabouts/safety-other-roundabout-benefits

40. Savolainen, P. T., Gates, T. J., Gupta, N., Megat-Johari, M.-U., Cai, Q., Imosemi, S., Ceifetz, A., McArthur, A., Hagel, E. C., & Smaglik, E. J. (2023, January 30). Evaluating the Performance and Safety Effectiveness of Roundabouts – An Update. Bts.gov. https://rosap.ntl.bts.gov/view/dot/68864 https://rosap.ntl.bts.gov/view/dot/68864

41. SEMCOG Crash Locations Map. (2026). SEMCOG.org. https://maps.semcog.org/CrashLocations/#/?andor=and&mcd_dcode=mcd&map_option=inj https://maps.semcog.org/CrashLocations/#/?andor=and&mcd_dcode=mcd&map_option=inj

42. Smith, T. (2024, September 3). Tough Love & Real Talk on Equity in Roadway Safety. Vision Zero Network. https://visionzeronetwork.org/real-talk-on-equity-in-roadway-safety/#:~:text=We%20won’t%20reach%20Vision,solutions%20to%20systemic%2Dlevel%20problems.https://visionzeronetwork.org/real-talk-on-equity-in-roadway-safety/#:~:text=We%20won’t%20reach%20Vision,solutions%20to%20systemic%2Dlevel%20problems

43. Smith, T. (2024, September 3). Tough Love & Real Talk on Equity in Roadway Safety. Vision Zero Network. https://visionzeronetwork.org/real-talk-on-equity-in-roadway-safety/?utm_source https://visionzeronetwork.org/real-talk-on-equity-in-roadway-safety/?utm_source

44. Stanton, R. (2023, June 2). These are the 10 wealthiest counties in Michigan, new study says. Mlive. https://www.mlive.com/news/2023/06/these-are-the-10-wealthiest-counties-in-michigan-new-study-says.html https://www.mlive.com/news/2023/06/these-are-the-10-wealthiest-counties-in-michigan-new-study-says.html

45. Steve. (2018, July 31). Michigan Roundabouts – Dangerous Intersections or a Step Toward Safety? Sinas Dramis Law Firm. https://sinasdramis.com/roundabouts-dangerous-intersections/https://sinasdramis.com/roundabouts-dangerous-intersections/

46. Tam, A. (2022, June 19). An Antiracist Future For Our Highways. Institute for Transportation and Development Policy - Promoting Sustainable and Equitable Transportation Worldwide. https://itdp.org/2022/06/19/an-antiracist-future-for-highways/https://itdp.org/2022/06/19/an-antiracist-future-for-highways/

47. Taylor, N. L., Porter, J. M., Bryan, S., Harmon, K. J., & Sandt, L. S. (2023). Structural Racism and Pedestrian Safety: Measuring the Association Between Historical Redlining and Contemporary Pedestrian Fatalities Across the United States, 2010‒2019. American Journal of Public Health, 113(4), 420–428. https://doi.org/10.2105/ajph.2022.307192 https://doi.org/10.2105/ajph.2022.307192

48. Tomer, A. (2011). Transit Access and Zero- Vehicle Households. https://www.brookings.edu/wp-content/uploads/2016/06/0818_transportation_tomer.pdf https://www.brookings.edu/wp-content/uploads/2016/06/0818_transportation_tomer.pdf

49. Traffic injuries much more common in poor areas, study finds. (n.d.-a). https://www.latimes.com/health/la-xpm-2012-apr-19-la-heb-road-crashes-poor-neighborhoods-20120419-story.html https://www.latimes.com/health/la-xpm-2012-apr-19-la-heb-road-crashes-poor-neighborhoods-20120419-story.html

50. Traffic injuries much more common in poor areas, study finds. (n.d.-b). https://www.latimes.com/health/la-xpm-2012-apr-19-la-heb-road-crashes-poor-neighborhoods-20120419-story.html https://www.latimes.com/health/la-xpm-2012-apr-19-la-heb-road-crashes-poor-neighborhoods-20120419-story.html

51. What is Vision Zero. (2026, January 6). Vision Zero Network. https://visionzeronetwork.org/about/what-is-vision-zero/https://visionzeronetwork.org/about/what-is-vision-zero/

Appendix A Roundabouts used in the Study

County	Year Built	Name	Coordinates
LIVINGSTON	2009	Hartland Rd./Rovey Dr.	42.63682, −83.74883
OAKLAND	1999	Tienken Rd./Runyon Rd./Washington Rd.	42.69743, −83.11135
OAKLAND	2001	Tienken Rd./Sheldon Rd.	42.69626, −83.12320
MACOMB	2003	Utica Rd./Dodge Park Rd.	42.59179, −83.01110
LIVINGSTON	2005	Lee Rd./US 23 NB Ramps/Fieldcrest Dr.	42.50486, −83.75652
LIVINGSTON	2006	Lee Rd./US 23 SB Ramps	42.50605, −83.75954
LIVINGSTON	2006	Lee Rd./Whitmore Lake Rd.	42.50627, −83.76012
OAKLAND	2004	Baldwin Rd./Indianwood Rd./Coats Rd.	42.78413, −83.31979
MACOMB	2006	25 Mile Rd./Hayes Rd.	42.70090, −82.97623
OAKLAND	2007	W. Maple Rd./Drake Rd.	42.54184, −83.39996
OAKLAND	2007	W. Maple Rd./Farmington Rd.	42.54241, −83.38035
OAKLAND	2008	Bogie Lake Rd./Cooley Lake Rd.	42.61317, −83.49239
LIVINGSTON	2000	Main St./S 3rd St.	42.52942, −83.78761
OAKLAND	2006	Loop Rd./Commerce Crossing Dr.	42.53542, −83.44801
OAKLAND	2008	Cooley Lake Rd./Oxbow Lake Rd.	42.61351, −83.49105
OAKLAND	2008	14 Mile Rd./Farmington Rd.	42.52785, −83.37922
OAKLAND	2009	Martin Pkwy./Library Dr.	42.55755, −83.44985
OAKLAND	2009	Martin Pkwy./PGA Dr.	42.56583, −83.44967
OAKLAND	2009	Martin Pkwy./Oakley Park Rd.	42.57048, −83.44984
OAKLAND	2015	W 14 Mile Rd./Orchard Lake Rd.	42.52846, −83.36020
OAKLAND	2008	Grand River Ave./New Hudson Dr.	42.51286, −83.62158
OAKLAND	2009	Grand River Ave./Lyon Center Dr. E	42.50967, −83.60697
OAKLAND	2010	New Hudson Dr./W Pontiac Trl.	42.50909, −83.62100
OAKLAND	2009	White Lake Rd./Duck Lake Rd. N	42.69157, −83.57226
OAKLAND	2009	White Lake Rd./Rose Center Rd. E	42.69312, −83.57431
OAKLAND	2010	W Hamlin Rd./S Livernois Rd.	42.65198, −83.15281
OAKLAND	2014	W Tienken Rd./N Livernois Rd.	42.69527, −83.15397
MACOMB	2009	26 Mile Rd./M-53 SB Ramps	42.71388, −83.02834
MACOMB	2009	26 Mile Rd./M-53 NB Ramps	42.71396, −83.02374
WAYNE	2014	Kercheval Ave./Wayburn St.	42.38076, −82.94229
OAKLAND	2005	Ryder Dr./Everett Dr.	42.63069, −83.11884
OAKLAND	2005	Ryder Dr./Connors Dr.	42.63060, −83.11736
OAKLAND	1994	Firewood Dr./Raintree Dr.	42.68744, −83.20810
OAKLAND	1994	Coachwood Ln./Falcon Dr.	42.69232, −83.20846
OAKLAND	2015	Pioneer Dr./Library Dr.	42.66886, −83.21576
MONROE	2013	St Anthony Rd./US 23 NB Ramps	41.79823, −83.68614
MONROE	2013	St Anthony Rd./US 23 SB Ramps	41.79802, −83.69021
LIVINGSTON	2017	Chilson Rd./E Coon Lake Rd.	42.53678, −83.87150
OAKLAND	2017	Napier Rd./W 10 Mile Rd.	42.46402, −83.55352
OAKLAND	2009	Taft Rd./Morgan Blvd.	42.44300, −83.49355
OAKLAND	2015	Civic Center Dr./Evergreen Rd.	42.48011, −83.24083
OAKLAND	2015	Evergreen Rd/(library and shopping center parking lots)	42.48345, −83.24100
MACOMB	2016	25 Mile Rd E/Romeo Plank Rd.	42.70099, −82.95338
MACOMB	2008	19 Mile Rd./Romeo Plank Rd.	42.61339, −82.93285
MACOMB	2010	Romeo Plank Rd./Canal Rd.	42.60022, −82.93216
MACOMB	2016	Jefferson Ave./Rosso Hwy.	42.63106, −82.82277
OAKLAND	2018	S Baldwin Rd./Gregory Rd.	42.72153, −83.30757
OAKLAND	2018	S Baldwin Rd./Judah Rd.	42.71519, −83.30749
LIVINGSTON	2008	Hamburg Rd./Winans Lake Rd.	42.46351, −83.79812
OAKLAND	2016	Crescent Blvd./Town Center Dr.	42.48561, −83.46982
St. CLAIR	2018	Lapeer Rd./Allen Rd.	42.98365, −82.52408
LIVINGSTON	2006	Village Place Blvd./Green Oak Ave.	42.50698, −83.75462
MACOMB	2005	Van Dyke Ave. (M-53) / 18 1/2 Mile Rd.	42.60148, −83.03119
LIVINGSTON	2013	Maltby Rd./(middle school access)	42.50039, −83.78169
MACOMB	2003	William Durant Blvd./Charles Chayne Rd.	42.50727, −83.04056
MACOMB	2004	Edward Cole Blvd./Louis Chevrolet Rd.	42.51705, −83.03027
WASHTENAW	2017	Nixon Rd./Dhu Varren Rd./Green Rd.	42.31702, −83.70779
WASHTENAW	2009	Huron Pkwy./Nixon Rd.	42.30516, −83.70748
WASHTENAW	2017	Scio Church Rd./S Wagner Rd.	42.25569, −83.79885
WASHTENAW	2018	Baker Rd./Dan Hoey Rd.	42.32693, −83.88729
WASHTENAW	2018	Baker Rd./Shield Rd./Dongara Dr.	42.32529, −83.88714
WASHTENAW	2016	Ann Arbor-Saline Rd./Textile Rd.	42.19859, −83.79691
WASHTENAW	2018	W Bemis Rd./Moon Rd.	42.17062, −83.73832
WASHTENAW	2016	Whittaker Rd./Merrit Rd.	42.18812, −83.60861
WASHTENAW	2015	Textile Rd./Hitchingham Rd.	42.20169, −83.62088
WASHTENAW	2016	Textile Rd./Stony Creek Rd.	42.20172, −83.62311
WASHTENAW	2010	Whittaker Rd./Stony Creek Rd.	42.20999, −83.61936
WASHTENAW	2017	Wiard Rd./Airport Dr.	42.23730, −83.56309
WASHTENAW	2013	Geddes Rd./Ridge Rd.	42.27679, −83.55419
WASHTENAW	2008	Geddes Rd./Superior Rd.	42.27492, −83.63755
WASHTENAW	2002	Campus Pkwy./Suncrest Dr./(parking)	42.18764, −83.74729
WASHTENAW	2002	Campus Pkwy./Community Dr.	42.18982, −83.74387
WASHTENAW	2007	Maple Rd./M-14 WB Ramps	42.30178, −83.78112
WASHTENAW	2007	Maple Rd./M-14 EB Ramps	42.29992, −83.78104
WASHTENAW	2007	Maple Rd./(Skyline High School Entrance)	42.30557, −83.78114
WASHTENAW	2013	State St./Ellsworth Rd.	42.22938, −83.73901
WASHTENAW	2011	Geddes Rd./US 23 NB Ramps	42.27440, −83.67437
WASHTENAW	2011	Geddes Rd./US 23 SB Ramps	42.27450, −83.67833
WASHTENAW	2011	Geddes Rd./Earhart Rd.	42.27460, −83.68216
WASHTENAW	2016	M-52 (Stockbridge Chelsea Rd.)/Werkner Rd.	42.33671, −84.02885
WASHTENAW	2017	W 8 Mile Rd./US 23 SB Ramps	42.42923, −83.76779
WASHTENAW	2017	W 8 Mile Rd./US 23 NB Ramps	42.42882, −83.76591
WASHTENAW	2017	N Territorial Rd./US 23 NB Ramps	42.37969, −83.75634
WASHTENAW	2017	N Territorial Rd./US 23 SB Ramps	42.37952, −83.75786
LIVINGSTON	2017	W 8 Mile Rd./Whitmore Lake Rd.	42.42944, −83.76903
OAKLAND	2018	E Flint St./Orion Rd./Miller Rd.	42.78398, −83.23184
OAKLAND	2018	N Squirrel Rd./Squirrel Ct.	42.63475, −83.22058
MACOMB	2006	Partridge Creek Blvd./Scoter Ln./Montage	42.62068, −82.94597
MACOMB	2006	Partridge Creek Blvd./Hawk Ln./Montage	42.62063, −82.94027
OAKLAND	2017	Franklin Rd./W Eleven Mile Rd.	42.48636, −83.29104
WAYNE	2005	Yellowstone Dr./Johnson Creek Dr.	42.40412, −83.52414
WAYNE	2004	Carriage Way/Glacier Rd.	42.40220, −83.53576
WAYNE	2004	Johnson Creek Dr./(park entrance)	42.39610, −83.51860
OAKLAND	2018	Pioneer Dr./Meadow Brook Rd.	42.66857, −83.21853
OAKLAND	2012	N Pontiac Trl./M-5 / Martin Pkwy.	42.55450, −83.44852
WASHTENAW	2007	E Ann St/Observatory St./(hospital access)	42.28240, −83.73114
LIVINGSTON	2008	Kensington Rd./Jacoby Rd.	42.55436, −83.69828
LIVINGSTON	2006	Village Place Blvd./Green Oak Dr./Fieldcrest Dr.	42.50573, −83.75296
OAKLAND	2017	Bell Rd./Coventry Woods Ln.	42.49306, −83.27073
MACOMB	2013	Metro Pkwy./Sail Rd.	42.57735, −82.79673
MACOMB	2018	33 Mile Rd./McKay Rd.	42.81797, −83.00023

Appendix B Figures

Note: For the following figures, the actual MEV values are listed below the graph figures. This is to make it easy to analyze numbers that aren’t counted as separate figures, but rather extensions of the current figures.

GitHub for further analysis of how we collected the figures: https://github.com/Haashir-Ahmed/Round-about-Analysis-MI

Figure 1.

This figure analyzes accidents that aren’t attributed to reckless behaviors. This information highlights a gap between roundabouts and non-roundabout intersection operations. Roundabouts in low-income communities experience these crashes at a rate of ~4.54 crashes per MEV, which is more than double the rate of crashes as seen in roundabouts in high-income communities. This difference demonstrates how operational challenges such as higher congestion, older vehicle maintenance, or less familiarity with roundabout navigation in lower-income areas lead to a higher frequency of minor, standard traffic incidents.

	Avg No Factor Rate
Income	Non-Roundabout	Roundabout
High Income	0.258774	2.484994
Low Income	0.329201	4.537415

	Average Distracted Rate
Income	Non-Roundabout	Roundabout
High Income	0.028281	0.141903
Low Income	0.027392	0.126246

Figure 2.

Distracted driving shows an unexpected result, with it being a more prevalent factor in roundabouts in high-income communities than in roundabouts in low income area. High-income communities show a distraction crash rate of ~0.14 MEV compared to ~0.13 MEV in low-income communities, while non-roundabout intersections show similar rates of crashes regardless of income. This figure demonstrates that distracted driving is not strongly correlated withcome-level.

	Average Drug Rate
Income	Non-Roundabout	Roundabout
High Income	0.003513	0.021480
Low Income	0.004551	0.020789

Figure 3.

Drug-related crashes are very rare overall and follow a similar counter-intuitive pattern to distraction, where roundabouts in high-income communities show a slightly higher crash rate, ~0.0214 MEV, compared to low-income communities, ~0.0207 MEV. This is the opposite of standard intersections, where low-income communities typically show higher drug involvement statistics. This demonstrates how roundabouts in wealthy areas still suffer impairment issues.

	Average Alcohol Rate
Income	Non-Roundabout	Roundabout
High Income	0.012883	0.073144
Low Income	0.017043	0.061149

Figure 4:

Alcohol involvement data present a deviation from standard traffic trends, in which low-income communities usually suffer from higher alcohol crash rates at standard intersections. Surprisingly, high-income communities have a rate of ~0.073 MEV compared to ~0.061 MEV in low-income communities. This ~19.67% increase demonstrates how roundabouts in high-income communities may be more prone to accidents, such as those in entertainment areas.

	Average Speeding Rate
Income	Non-Roundabout	Roundabout
High Income	0.025182	0.208781
Low Income	0.033581	0.093604

Figure 5.

This figure shows speeding, which represents one of the most significant behavioral deteriorations in wealthy areas. This is seen with high-income area roundabouts experiencing a speeding crash of ~0.209 MEV, which is ~123.05% higher than the ~0.094 MEV rate found in roundabouts in low-income communities. This is a sharp contrast to standard intersections, where the low-income communities had higher speeding-related crashes.

	Average Hit&Run Rate
Income	Non-Roundabout	Roundabout
High Income	0.036474	0.335626
Low Income	0.124199	0.693718

Figure 6.

Hit-and-runs show the most dramatic socioeconomic disparity of all factors. Roundabouts in low-income communities suffer from a hit-and-run rate of ~0.694 MEV, which is roughly double the high-income area roundabout rate of ~0.336 MEV. This gap is even greater in similar intersections where low-income hit-and-run rates are more than triple those of high income area. In higher-income areas, this gap is even wider, with hit-and-run rates almost ten times higher for roundabouts than at similar intersections.