October 13, 2022 · 11:00 am - 12:00 pmNovember 18, 2022 · 9:00 am - 12:30 pmOctober 6, 2022 2022 Speaker Series on California's Future — Virtual EventOctober 4, 2022 2022 Speaker Series on California's Future — Virtual EventWe believe in the power of good information to build a brighter future for California.Help support our mission.Magnus Lofstrom, Joseph Hayes, Brandon Martin, and Deepak PremkumarSupported with funding from Arnold VenturesStark racial inequity has long been a deeply troubling aspect of our criminal justice system.In recent years, traffic stops have emerged as a key factor driving some of these inequities and an area of ​​potential reform.Are there opportunities to identify kinds of traffic stops that could be enforced in alternative ways—potentially improving officer and civilian safety, enhancing police efficiency, and reducing racial disparities—without jeopardizing road safety?To explore this question, in this report we use data on 3.4 million traffic stops made in 2019 by California's 15 largest law enforcement agencies to examine racial disparities in stop outcomes and experiences across time of the day, type of law enforcement agency, and type of traffic violation.These findings suggest that nighttime traffic stops for non-moving violations—especially those made by police and sheriff departments—deserve consideration for alternative enforcement strategies.However, any changes need to be balanced against the possibility of hampering efforts to confiscate dangerous contraband, especially firearms.Evidence-based practices to reduce racial bias, such as diversifying policing staff, could also help mitigate inequities and improve public safety.In 2020, the killing of George Floyd by a Minneapolis police officer—and the deaths of other African Americans in police custody—sparked widespread protests and intensified concerns about persistent racial inequities in the criminal justice system.In recent PPIC surveys, 61 percent of California adults believe that the criminal justice system is biased against Black individuals, with 84 percent of Black Californians holding this view (Baldassare et al. 2022).And while 54 percent of California adults say police treat all racial and ethnic minorities fairly “almost always” or “most of the time,” only 18 percent of Black Californians share this opinion (Lawler and Thomas 2021).Efforts to reduce racial inequities in people's experiences with police are critical to improving community engagement and trust in law enforcement.To begin answering questions about the size of these racial disparities, the contexts in which they occur, and the groups of people being affected, California passed the Racial and Identity Profiling Act (RIPA) in 2015. Implementation has been rolled out in waves based on the size of the agency.Since 2018, the state's largest law enforcement agencies have been collecting officer-perceived data on demographics and other details for all pedestrian and traffic stops, and reporting this information to the California Department of Justice, with more agencies required to report each year.By 2023, all law enforcement agencies in California will be required to collect and submit stop data.This rich data source has already yielded important insights into racial disparities in police stops and encounters.Recent research finds that Black Californians are more than twice as likely to be searched as white Californians, but searches of Black Californians are somewhat less likely to yield contraband or evidence.While differences in jurisdiction and context significantly contribute to racial disparities in experiences with law enforcement officers, notable inequities remain even after accounting for such factors (Lofstrom et al. 2021).This research also identifies that disparities—especially between Black and white individuals—are greatest in traffic stops made by local law enforcement (ie, police and sheriff departments, as opposed to the California Highway Patrol).Interactions between civilians and police, whether for a traffic stop or another reason, carry inherent risks.In 2019, 799 assaults on police officers occurred during traffic pursuits and stops in California (7.6% of total assaults on law enforcement)—55 of which involved a firearm (19% of all firearm-related assaults on law enforcement).In addition, in a study of police use of force, Premkumar et al.(2021) finds that traffic and pedestrian stops account for about 15 percent of police encounters in which a civilian is seriously injured or killed.Overall, there are stark racial disparities in civilians injured during law enforcement encounters.Black Californians are about three times more likely to be seriously injured, shot, or killed by the police relative to their share of the state's population.Concerns about these racial disparities in the use of force have motivated several statewide policing reforms.In addition, a few cities in California, including Los Angeles, San Francisco and Berkeley, have proposed reforming enforcement of some traffic violations.In this report, we build on our previous work by using RIPA data to help identify traffic stops that may deserve consideration for alternative enforcement practices.Traffic stops that could be enforced using alternative methods would be those that are unlikely to jeopardize public or road safety but that could: (1) improve safety for officers and civilians, (2) increase police efficiency, and (3) reduce racial disparities.Since agencies have substantial differences in their primary missions, objectives, and jurisdictions, we examine stop outcomes by type of law enforcement agency (the California Highway Patrol or local law enforcement; the latter can be further separated into police and sheriff departments).Furthermore, we analyze traffic stops for both moving violations (eg, speeding or failure to stop) and non-moving violations (eg, improper display of a license plate, expired registration tag, or failure to maintain vehicle light equipment).Non-moving violations in particular may be an area where alternative enforcement—such as mailing the vehicle owner a “fix-it ticket” and/or citation, as proposed by San Francisco Police Chief Bill Scott—is feasible and safe.Traffic stops that do not lead to any enforcement or discovery of contraband or evidence deserve special attention—while these incidents are a small minority of all stops, they fall disproportionately on people of color and may not be an efficient use of law enforcement officers' time (Lofstrom et al. 2021).The “intrusiveness” of these stops, as measured by outcomes such as whether individuals were asked to step out of the vehicle, searched, detained, or handcuffed, and whether an officer aimed or used a weapon, is another critical factor that could affect the relationship between community members and law enforcement.Importantly, we also examine the extent to which stop outcomes and racial disparities vary throughout the day.Such information can be used to determine when changes in policing practices may be warranted and could help guide potential changes.If, for example, searches during certain times of the day are particularly unlikely to yield contraband or evidence, this may present an opportunity to redirect such efforts and resources to other policing tasks.One key question is whether racial disparities in traffic stops are the result of bias or targeting on the part of law enforcement.While disparities could be driven by racial bias, other factors may also play a role.For example, vehicle condition as well as driving patterns and behavior may differ across race/ethnicity and could lead to disparities in the likelihood of being stopped for a traffic violation.To examine the potential role of racial bias in traffic stops, we employ the “veil of darkness” theory, which posits that it is more difficult for officers to ascertain a person's race or ethnicity during dark hours.If officers are racially profiling drivers, the share of people of color in traffic stops would be lower during dark hours, compared to light hours, holding everything else constant.Using the shift to and from Daylight Saving Time, we examine if people of color are more or less likely to be stopped for a traffic violation when, for a given time of the day, sudden changes in light conditions make it easier or harder, respectively , to determine an individual's race or ethnicity.Note that the test of the “veil of darkness” theory only applies to the likelihood of being stopped for a traffic violation and cannot be applied to other stop outcomes, as once a stop has been made, the officer has determined (the perceived) race /ethnicity of the driver.Furthermore, while nighttime conditions make it more difficult to determine a driver's race/ethnicity than during daytime, there are contexts and locations, such as where there are streetlights, when racial/ethnic identification may still be possible.Officers may also use type, make, and condition of a vehicle to infer race/ethnicity.If such factors are indeed correlated with race/ethnicity, racial profiling may also be possible during dark hours.This report begins with an overview of the RIPA data on police stops.Using data from 2019 for the state's 15 largest law enforcement agencies, we then examine traffic stops across different times of the day by agency, type of traffic violation, and intrusiveness.Next, we focus on racial disparities in traffic stops across these same dimensions.We then investigate the potential role of racial bias in traffic stops.Finally, we conclude with additional considerations for policy and practice based on this research.Recognizing the need for data and research on law enforcement stops, California passed the Racial and Identity Profiling Act in 2015 (Assembly Bill 953) in an effort to directly confront concerns over racial inequities in policing.RIPA requires all law enforcement agencies in California to collect officer-perceived demographic data and other detailed data regarding all pedestrian and traffic stops by 2023. A “stop” is defined as any detention by a peace officer of a person, or any peace officer interaction with a person in which the officer conducts a search.The data elements mandated by statute include individual-level and stop-level information.For individual-level data, officers are required to record their perception of the identity characteristics for each person stopped, including race or ethnicity.Other perceived traits included in the RIPA data are gender, approximate age, LGBT (lesbian, gay, bisexual, or transgender) status, English fluency, and disability (including behavioral health status).Officers are prohibited from asking the person stopped to self-identify these characteristics.The data do not allow for corroborating the accuracy of the reported information, including the race and identity of the individual stopped and the specific actions taken by the officer.Nor do the data include information on the race and ethnicity of the officer.In this report, we rely on 2019 RIPA data.While 2020 RIPA stop data are now available, 2020 was a highly unusual year due to the COVID-19 pandemic and the numerous public health challenges it presented.These challenges led to many changes and disruptions as our criminal justice systems aimed to reduce the spread of the coronavirus.It was also a year of civil unrest and racial reckoning following the killing of George Floyd.In addition, many parts of the state experienced significant swings in crime rates (Lofstrom and Martin 2022).All of these factors raise the concern that police stops reported in the 2020 RIPA data may reflect these extraordinary circumstances rather than longer-term patterns in traffic enforcement.The 2019 RIPA data includes all traffic stops (3,394,392 stops, about 85% of all 2019 stops) reported by the 15 largest law enforcement agencies in the state.This includes the California Highway Patrol (CHP), eight police departments (Los Angeles, San Diego, San Francisco, Sacramento, Fresno, San Jose, Long Beach, and Oakland), and six county sheriff's departments (Los Angeles, San Bernardino, Sacramento, San Diego, Riverside, and Orange County).Given that the analysis is limited to large law enforcement agencies, of which the Oakland Police Department is the smallest with 740 sworn officers, the findings do not speak to experiences during traffic stops in the large number of small law enforcement agencies throughout the state.For more details about the data, including detailed descriptive statistics, see Technical Appendix A in Lofstrom et al.(2021).Throughout this report, we frequently separate stop characteristics by type of law enforcement agency.The CHP occupies a unique place in this group, with its jurisdiction over the entire geography of the state and its mandate to focus on enforcing traffic laws.The nature of this mandate, and the scope of the CHP's operations, are reflected in the sheer number of stops it makes, as well as the reasons behind these stops.Sheriff departments, with jurisdiction over their respective counties, and municipal police departments, responsible for their own incorporated cities, also enforce traffic laws but have a wider mandate to uphold public safety.The jurisdictions for sheriff and police departments vary considerably in geography, the public safety environment, and the size, density, and racial/ethnic composition of the population.In this report, we occasionally report sheriff and police departments' results separately when doing so helps illustrate the differences between their operations;other times, we combine these categories under the term “local” law enforcement agencies to draw a contrast with findings from the CHP.We begin our analysis of the nearly 3.4 million traffic stops made in 2019 by these 15 large law enforcement agencies (LEAs) by examining how traffic patterns affect stops and stop outcomes.Unsurprisingly, traffic stops peak during prime commuting hours, when there are the most cars on the road.As Figure 1 shows, roughly a fifth of daily stops take place between 7 am and 10 am, and nearly another fifth occur between 3 pm and 6 pm Notably, a large number of traffic stops also take place in the hours before and after midnight: about a fifth of traffic stops in a typical day, between 80,000 and 140,000 stops per hour, take place between 9 pm and 3 am This relatively high share of traffic stops is notable considering that there are far fewer drivers on the road during these hours.SOURCE: Author calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTE: Stop hours are shown using the 24-hour clock, sometimes referred to as “military time,” where 0 hour represents midnight.Consistent with its organizational mission, the CHP makes many more traffic stops than the police or sheriff departments.Overall, the CHP made 64 percent of the nearly 3.4 million traffic stops (almost 2.2 million), while the eight police departments made 24 percent of these stops (about 809,000) and the six sheriff departments made about 12 percent of these stops (about 425,000 ).Slightly more than one-third of all traffic stops are for non-moving violations.Examples of some of the most common non-moving violations are lack of registration (287,900 stops), improper display of a license plate (167,800), and failure to maintain vehicle light equipment (67,100).Figure 2 shows the shares of stops made for non-moving violations throughout the day, broken down by type of law enforcement agency.Notably, in the hours before and after midnight, local law enforcement agencies are especially likely to make stops for non-moving violations.Between 9 pm and 3 am, the share of stops for non-moving violations rises to roughly half of stops made by police departments and almost two-thirds of stops made by sheriff departments, while this share drops to about a fourth of all CHP traffic stops.In contrast, during most regular work and business hours, the shares of stops made for non-moving violations tend to be somewhat lower for local law enforcement agencies than for the CHP.SOURCE: Author calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTE: Stop hours are shown using the 24-hour clock, sometimes referred to as “military time,” where 0 hour represents midnight.The likelihood of being searched during a traffic stop, or the search rate, varies notably throughout the day, from about one out of every hundred traffic stops during the peak morning commuting hours to about one out of every ten traffic stops from 8 pm to 3 a.mHowever, discovery rates—that is, how often searches yield contraband or evidence—in stops during the hours before midnight are overall no higher than at other times of the day, and discovery rates are notably lower during the hours after midnight.In fact, despite a relatively higher likelihood of being searched in the hours after midnight, searches made at this time are the least likely to yield contraband or evidence.Officers discover contraband or evidence in about 15 percent of searches in the hours after midnight, compared with 18–22 percent in searches during all other hours.This, however, differs quite widely across agencies and race/ethnicity (see Figure A1 in Technical Appendix A).However, a number of guns are confiscated in traffic stops in the hours before and after midnight.In 2019, these 15 law enforcement agencies made 905 traffic stops that resulted in officers confiscating firearms;47 percent (421) of these stops took place between 9 pm and 3 am Of these stops, 51 percent (463) were for non-moving violations, at any time of day (320 by police officers, 98 by sheriff deputies, and 45 by the CHP).Altogether, traffic stops involving seizure of a firearm represented about 30 percent of the 3,024 stops of any kind in which a firearm was seized by these law enforcement agencies in 2019.In the overwhelming majority of stops, there is at least some enforcement, measured here as at least a warning being issued.Traffic stops that lead to no enforcement and that do not yield any contraband or evidence are arguably the types of stops most deserving of closer examination, as they could potentially be considered for alternative enforcement methods without endangering public safety.Of the close to 3.4 million traffic stops made by these 15 law enforcement agencies in 2019, only about 6 percent (about 211,000 stops) led to no enforcement or discovery of any contraband or evidence.But there is substantial variation depending on the time of day.For example, although stops made between 6 pm and 1 am represent only about a third of overall traffic stops, more than half of no-enforcement/no-discovery traffic stops take place between those hours.Figure 3 shows that stops by local law enforcement officers, especially police officers, disproportionately result in no enforcement or discovery.The CHP makes only 15 percent of traffic stops that result in no enforcement or discovery, despite making the majority of stops overall (64%).While the eight police departments account for slightly less than a fourth of overall traffic stops (24%), they make more than half of the traffic stops with no enforcement or discovery (59%).The six sheriff departments make about 13 percent of overall traffic stops and 26 percent of those with no enforcement or discovery.For local law enforcement, these stops are more likely to be for non-moving violations rather than moving violations.SOURCES: Authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTES: No enforcement is defined as not even a warning was issued.The figure shows the shares of moving and non-moving violation traffic stops made by each type of law enforcement agency in which there is no enforcement/no discovery of contraband or evidence (left three bars) and among all traffic stops (right three bars) .Lastly, as shown in Table 1, stops with no enforcement or discovery comprise a meaningful amount of officers' (and civilians') time.On average, these stops last about 23 minutes, but this is skewed by a relatively small share (about 5%) of stops in which the officer reports a duration longer than an hour.The median duration for this kind of traffic stop is 10 minutes.In addition, the length of the stop varies by agency type and to some extent the type of traffic violation.Notably, officers spent more than 80,000 total hours in 2019 on these types of stops (or 7% of total officer hours spent on traffic stops).For police departments, close to 28,000 of these hours were stops made for non-moving violations.SOURCES: Authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTE: No enforcement is defined as not even a warning was issued.The interaction between a civilian and an officer during a traffic stop can vary considerably.In addition to searches, officer actions can range from verbal communication to the civilian being asked to step out of the vehicle to the civilian being detained or handcuffed to—in rare cases—an officer aiming or using a weapon.We refer to this wide range of potential officer behaviors that extend beyond verbal communication as “intrusive actions.”While the majority of no-enforcement/no-discovery traffic stops entail only verbal communication with the officer, a number of these stops include intrusive actions, as shown in Table 2. For example, in 24 percent of these stops (nearly 50,000), the individual was asked to step out of the vehicle.The individual and/or vehicle was searched for contraband or evidence in about 18 percent of these stops (almost 37,400), and the person was detained curbside or in a patrol car during 17 percent of these stops (about 35,700).In about 7 percent of no-enforcement/no-discovery stops (nearly 15,000), the person was temporarily handcuffed and then released.Out of the roughly 211,000 no-enforcement/no-discovery stops, an officer aimed a weapon in 361 of these stops (0.1%) and used a weapon in 33 of these stops (0.02%).SOURCES: Authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTE: No enforcement is defined as not even a warning was issued.These patterns differ across law enforcement agencies and types of violations.Table 2 shows that regardless of level, intrusive experiences are most common in stops for non-moving violations made by local law enforcement agencies.For example, in the nearly 15,000 instances in which the officer reported handcuffing the individual without any further enforcement or discovery, almost half were in stops for a non-moving violation made by the eight police departments.In sum, the 2019 RIPA data show that while traffic stops peak during commuting hours, many stops take place during the late pm and early am Searches for contraband and evidence are most likely to occur during these hours, while discovery of them are least likely in the early am We also find that in about 6 percent of traffic stops there is no enforcement or discovery of contraband or evidence.About half of these stops take place in the late pm and early am, and almost 60 percent of these stops are made by the eight police departments—the majority for non-moving violations.During some of these no-enforcement/no-discovery stops, officers also report intrusive actions.For example, in about a quarter of these stops, the individual was asked to step out of the vehicle and in almost one in five, the individual and/or vehicle was searched for contraband or evidence.A key motivation for this report is to identify traffic stops that can be targeted for changes in enforcement practices that could reduce racial disparities—possibly improving police efficiency and community relations without jeopardizing public and road safety.In this section, we examine overall racial disparities in traffic stops and whether racial disparities are especially large in stops that occur during certain times of the day or for certain types of traffic violations.On average, 40 percent of traffic stops are of Latino drivers, 33 percent are of white drivers, and 14 percent are of Black drivers.With about 6 percent of the state's population, Black Californians are markedly overrepresented in traffic stops, as shown in Figure 4. Meanwhile, the share of traffic stops of Latinos is similar to their share of the population (40% of stops vs. 39% of the population), and white drivers are somewhat underrepresented in traffic stops (33% of stops vs. 35% of the population).At 6 percent of traffic stops, California's growing Asian population, now at 15 percent, is notably underrepresented.SOURCES: Stop share are authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019. Population data comes from the California Department of Finance.The racial composition of drivers stopped for traffic violations varies considerably throughout the day.The share of Black drivers changes most notably, from about 11 percent mid-morning to about 18–19 percent in the late pm and early am hours.Depending on the time of day, Black drivers are between two and almost four times as likely to be stopped relative to their share of the population.As Figure 5 shows, the racial disparity is especially stark in traffic stops made by police departments, where Black drivers account for twice the share of white drivers in the hours before and after midnight (29%–34% vs. 15%–17% ), despite making up a far smaller share of the population.As discussed further below in our “veil of darkness” analysis, these disparities could be driven by racial bias, but other factors may also play a role.Vehicle conditions, driving patterns, and driving behaviors may differ across race/ethnicity, and this racial variation could contribute to disparities in the likelihood of being stopped for a traffic violation at a given time of day.SOURCE: Authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTES: Stop hours are shown using the 24-hour clock, sometimes referred to as “military time,” where 0 hour represents midnight.Not all racial/ethnic groups in the RIPA data are shown in the figure, thus each bar will not add up to 100 percent.The racial/ethnic groups omitted in the figure are individuals perceived by the officer to be Middle East/South Asian, Pacific Islander, Native American, or multi-racial/ethnic.The likelihood of being searched during a traffic stop varies across race and ethnicity as well as across agency type.Black drivers stopped by local police and sheriff departments are searched in 20 percent of traffic stops, while the search rates for Latino and white drivers are 13 percent and 6 percent, respectively.The likelihood of being searched, and racial disparities therein, are noticeably smaller in traffic stops made by the CHP.On average, about 1.9 percent of Latinos stopped for a traffic violation by the CHP are searched for contraband, compared to 1.5 percent and 1.2 percent of Black and white drivers, respectively.Disparities in search rates vary noticeably throughout the day for local law enforcement stops, but not markedly for CHP stops, as shown by the solid lines in Figure 6 (numbers closer to zero reflect lower disparities at those times of day).The disparities in local law enforcement stops are especially stark between 4 pm and midnight.While roughly one in ten white drivers stopped by local law enforcement in the late evening are searched for contraband or evidence, about one in four Black drivers and one in five Latino drivers are searched.Meanwhile, the somewhat higher search rate of Latinos compared to white drivers in CHP traffic stops is similarly driven by stops made in the late pm and early amThe higher search rates of Latino and Black drivers in traffic stops made by local law enforcement are not associated with higher rates of discovery of contraband or evidence.In fact, discovery rates are markedly lower for Black and Latino drivers (at about 19% for both) than for white drivers (25%)—as shown by the negative numbers for the dotted lines in the top panel of Figure 6, which compares the discovery (or “hit”) rates for Black and Latino drivers with those of white drivers.In searches in traffic stops made by the CHP, disparities are negligible during most hours of the day, and disparities in discovery rates are inconsistent throughout the day.Overall, the detection rates in CHP stops are 13 percent, 12 percent, and 9 percent for white, black, and Latino drivers, respectively.As the bottom panel in Figure 6 shows, for CHP stops, the disparities in discovery rates fluctuate throughout the day, with higher discovery rates in searches of people of color during certain daytime hours.Note, however, that the likelihood that a CHP officer will conduct a search at this time of the day is very low, in only about 0.2–0.3 percent of stops, regardless of race/ethnicity.The discovery rates in these searches are markedly higher (between 20% and 40%) in these relatively rare searches, suggesting that they largely take place when the officer sees or smells contraband or evidence and/or that a higher threshold for engaging in a search results in more efficient discovery of contraband/evidence.SOURCE: Authors' calculations using California Department of Justice, Racial and Identity Profiling Act (RIPA) Wave 2 data, 2019.NOTES: Figures show percentage point differences in search and discovery rates between Black and white individuals and white and Latinos.Stop hours are shown using the 24-hour clock, sometimes referred to as “military time,” where 0 hour represents midnight.For the CHP, the shares of no-enforcement/no-discovery stops of white, Black, and Latino drivers do not differ much (about 1.4%–1.5% for all three groups).However, these shares do vary significantly in local law enforcement stops, from 11 percent for white drivers to 15 percent and 22 percent, respectively, for Latino and Black drivers.As Figure 7 makes clear, racial disparities in stops with no enforcement or discovery made by local law enforcement also fluctuate considerably throughout the day and are especially stark in the late pm and early am hours.