Introduction

Societal discussions surrounding the efficacy of police discretionary traffic stops arose in the mid-1990s with the “driving while Black” movement and have recently re-energized in the aftermath of George Floyd’s murder in the Summer of 2020 (Benet & McMillan, 2021). Since traffic stops are the most frequent type of police-citizen interaction (Bureau of Justice Statistics, 2018), an emerging policy recently adopted by one state (e.g., Virginia) and local agencies around the United States (e.g., Los Angeles, Philadelphia, Seattle) is to restrict discretionary traffic stops. That is, these jurisdictions limit stops for vehicle equipment-related infractions (e.g., broken taillights). Equipment-related stops are often used as a discretionary pretext to further investigate the driver/passengers for criminal behavior (Holder, 2023).

These discretionary stops have been argued to disproportionately target impoverished, minority drivers who may be unable to repair equipment-related vehicular issues (Baumgartner et al., 2018). Beyond equipment-related stops, discretionary stops in general have been suggested to be disproportionately used against minority drivers, potentially resulting in other adverse outcomes (e.g., arrest, use of force) (Smith, 2023). Although there is some research to support claims of racial disparity in discretionary stops (Roh & Robinson, 2009), other studies have not found this to be the case (see discussion in McCabe et al., 2021). Nonetheless, there has been an extensive call to restrict discretionary stops in a post-Floyd reform era of policing (Woods, 2021).

Discretionary traffic stops are a common and consistent function of police duties and are intended to exert presence in the community, deter and apprehend criminals, and check for outstanding warrants (Wu & Lum, 2019). Discretionary stops are an important aspect of police preventive patrols, allowing officers to use reasonable suspicion to stop and investigate civilians, potentially preventing jurisdictional crime rates. Scholars have found a link between proactive discretionary stops and crime reductions across jurisdictions (see discussion in Lum et al., 2020). However, following the passage of policies to restrict discretionary traffic stops, police have expressed concern that these restrictions may result in elevated jurisdictional crime rates (Lee, 2022). While activists and police have differing opinions on discretionary stops, with police arguing for their utility and activists pushing for reform, there are limited studies that experimentally assess not only the perceptions of the public who are most impacted by the stops but also the potential recipients of any crime control benefits/costs. Importantly, the absence of such studies may impact our understanding of how the public responds to these practices, consequently impeding the development of informed and effective policy reforms. The present study implements an original survey experiment assessing the impact of discretionary traffic stop data provision on public perceptions of these stops as an effective crime-fighting tool. Before presenting the study, we highlight relevant research.

The utility and controversy over discretionary traffic stops

Since landmark Supreme Court decisions in Terry v. Ohio (1968) and Whren v. United States (1996) allowed greater police discretion in executing stops of civilians, much scrutiny and research have emerged about the effectiveness of, and, potential disparities that may result from various police actions (Chin & Vernon, 2014; Dixon et al., 2008; Donahoe, 1996). Further, the “driving while Black” movement in the 1990s brought about increased scrutiny and controversy surrounding police traffic stop behavior and the outcomes of traffic stops (Harris, 1996). Some research has found racial disparities in police traffic stops, both in the disparity of who is stopped (Miller, 2008) and the outcome of traffic stops (Baumgartner et al., 2017; Pierson et al., 2020).

Despite the above evidence, some studies do not find such racial disparities in traffic stops and outcomes (Grogger & Ridgeway, 2006; Worden et al., 2012). For example, McCabe and colleagues (2021), using observational methods, found that in only one of the nine street segments analyzed, there were differences in the number of Black motorists stopped compared to White motorists. Alpert and colleagues (2004) examined demographic data of the not-at-fault parties of collisions involving two vehicles in selected intersections, and found non-significant racial differences in police traffic stop data at those same intersections. Thus, although there are consistent claims of racial disparity (and discrimination) in police traffic stop behavior, the empirical evidence is mixed.

Police discretionary traffic stops are common and core tools police use to remain present and visible within jurisdictions while providing a deterrent mechanism to keep crime levels at bay (see in-depth discussion in Wu & Lum, 2019). Not only may these stops impact offender’s decision calculus, but officers proactively checking licenses for outstanding arrest warrants, searching drivers/vehicles for dangerous contraband, and arresting civilians may contribute to crime control (see discussion in Epp et al., 2014). Some evidence supports these claims, particularly when discretionary traffic stops are executed in high-crime hot spots (Braga et al., 2019). Although the effect of traffic stops across larger geographic areas may not find as much support as micro-locations (although see Petersen and colleagues (2023), there is a body of literature showing an impact of discretionary traffic stops on crime (Boehme & Mourtgos, 2024; Nix et al., 2024). Moreover, police argue that these stops are necessary for executing daily duties and maintaining low crime rates (Epp et al., 2014; Lee, 2022). However, little is known about public approval of such strategies and whether data information provided to the public affects their support of these activities.

Empirical research of information provision on public opinion of criminal justice topics

Scholars have assessed the impact of information provision on public opinion on various criminal justice topics. For instance, scholars have experimentally presented various information to respondents to assess public approval of capital punishment (Norris & Mullinix, 2020; S. Wu, 2021; Suiter & Metcalfe, 2023), and information on the race and gender of the defendant and victim as to whether the public would label the offender as a sex offender (Stevenson et al., 2009). In the realm of policing, scholars have experimentally tested the impact of information on various policing outcomes such as police use of force (Mullinix et al., 2021), police effectiveness (Boehme et al., 2023), body-worn cameras (Demir, 2019), and police reform (Vaughn et al., 2022).

In particular, Boudreau and colleagues (2019) proposed that providing information outlining a trend of police violence shapes respondents’ attributions of blame and levels of trust in the police, with these effects contingent on their sociodemographic characteristics (see also Schiff et al., 2022). Vaughn and colleagues (2022) found that the public largely did not support abolishing or defunding the police movements after being provided various definitions of each movement. Of most relevance to the present study, Mullinix and Norris (2019) found that presenting objective information about racial disparities in pulled-over rates diminished respondents’ trust in police, particularly influenced by their beliefs about the primary cause of such disparities (also see other policing experiments in Mummolo, 2018a, 2018b; Nix et al., 2021). While there are certainly many more relevant studies, this research shows that public perceptions about the criminal justice system may be shaped by information on relevant topics. However, no known study has assessed the impact of data and information on discretionary traffic stops on the public’s approval of these stops as a crime-fighting tool.

Present study

In the aftermath of George Floyd’s death in 2020, several policies were proposed to reduce disparities and/or negative outcomes for civilians, particularly civilians of color (Boudreau et al., 2022). One such policy, which has been passed across the state of Virginia as well as local agencies across the United States (e.g., Los Angeles PD, Philadelphia PD, and Seattle PD), is the effort to restrict discretionary traffic stops (Boehme & Mourtgos, 2024). While we have heard from reform activists, police administrators and officers, and scholars regarding the utility and efficacy of discretionary traffic stops (Conklin & Pagones, 2022; Lee, 2022), little is known about the public perceptions of these stops. Further, it is yet known whether data provision on the outcomes of discretionary stops shapes such perceptions on the effectiveness of these stops on crime control/prevention.

The present study seeks to assess whether randomly presenting the public with data on discretionary traffic stops causally impacts respondents’ perceptions of the crime prevention effects of these stops. To assess whether data impacts support for the use of discretionary stops, particularly during a time when the utility of these stops of under great scrutiny, we randomly present respondents with one of two sets of data: contraband hit rates or racial disparities of traffic stops. Results from this study may provide important implications for police, local governments, and understanding public perceptions. For example, this study may highlight that an informed public through data may alter perceptions about routine policing strategies, and an all-out ban on traffic stops is not only unsupported by police but also the public. Additionally, understanding which data moves public opinion on a commonly used strategy will inform information sharing from police agencies to the public. Such information sharing and transparency can lend legitimacy to an agency and, in turn, increase compliance with the law by members of the community (President’s Task Force on 21st Century Policing, 2015). These improved relations could help improve officer morale if they have the support of the public in using discretion during traffic stops. These findings are particularly timely considering various jurisdictions nationally are considering adopting restrictive policies about discretionary stops. Below we seek to test the following hypotheses.

  • H1: In comparison to the control condition, those randomly assigned the “contraband seizure” condition are more likely to support the use of discretionary police stops.

  • H2: In comparison to the control condition, those randomly assigned the “disparity” condition are less likely to support the use of discretionary police stops.

Methods

We distributed an original survey experiment to all heads of households across the state of South Carolina with a listed email address (N = 904,531) to test our hypotheses. The list of addresses was obtained from Mailer’s Haven, a third-party listserv (Mailers Haven, 2023). Although there are concerns over representativeness of our probability sample that contains only heads of households with an accompanying email address, there is evidence that our sub-sample may still elicit externally valid results (Keeter & McGeeney, 2015; Patten & Perrin, 2015). We used a dynamic distribution and recruitment procedure to collect a racial/ethnically diverse sample of respondents. We had a diverse research team involved in respondent recruitment, including researchers from Asian, Black, Hispanic, and White racial/ethnic backgrounds. Therefore, email invitations to the Asian sample would be sent from a member of the research team of Asian descent, a Black member of the research team would distribute emails to the Black sample, and so on, which is an empirically supported approach in recruiting diverse samples and populations (see Yancey et al., 2006). Although emails may have been distributed by various members of the research team, the recruitment messages across the various lists were identical. At the conclusion of data collection, we obtained a sample containing about 1.2% Asian, 14.7% Black, 2.8% Hispanic, and 76.2% White, and the remaining as Other (detailed below).

The survey was distributed via Qualtrics beginning on October 13, 2023, with five periodic email reminders throughout with data collection ending on December 13, 2023. After data cleaning and survey dropout, we obtained a final sample of 4740 respondents for analysis.Footnote 1 We rely on the American Association of Public Opinion Research (AAPOR, 2023) RR2 estimates to calculate our sample size which suggested a 0.05% response rate; however, we are unable to know the true number of respondents that actually received the survey in their inbox (e.g., not filtered into junk/spam), suggesting our true response rate is likely higher than the RR2 estimate.Footnote 2 Further, there is evidence to suggest that nonresponse does not lead to biased estimates (Pickett et al., 2018). Even with this evidence, we are cautious in extrapolating our results broadly across the United States.

The survey was developed by a cross-University and cross-Department research collaboration of faculty and graduate students. The survey asked respondents about various related topics including governmental responses to crime problems, including drug crimes/use, and drug overdose scenarios. The survey was pilot tested on three separate occasions by faculty, post-doctorate fellows, and graduate students from various departments and universities. This study was approved by the hosting University’s Institutional Review Board (Protocol #Pro00131209). For the present study, we report findings from an experimental portion of the study that randomly assigned respondents into one of three conditions: (1) contraband seizure condition (hereafter “contraband condition/treatment”), (2) racial disparity condition (hereafter “disparity condition/treatment”), or (3) the control condition. If respondents were randomly assigned to a treatment, they would receive data information derived from one study that investigated traffic stop outcomes on one page.Footnote 3 Then, respondents would have to click to the next page where they would again see the same information just above a list of Likert-scale matrix questions. Respondents receiving the information twice were purposeful to ensure that respondents were aware of the treatment. If receiving the control condition, respondents were presented with the same Likert-scale matrix questions but did not receive any data information. The treatments appeared verbatim as follows (presented in bold font to the respondent):

Contraband condition

One study that analyzed over 18 million police traffic stops found that when police pull over a vehicle and search the vehicle/driver/passenger(s), police found and confiscated contraband (e.g., drugs, firearms) in 25% of these searches.

Disparity condition

One study that analyzed over 18 million police traffic stops found that Black drivers were 55% more likely to be searched during a traffic stop compared to White drivers, although contraband (e.g., drugs, firearms) was less likely to be found on Black drivers.

Variables

Our dependent variable is a summed scale of respondents’ answers of how much they agree or disagree with the following four statements about police discretionary traffic stops (1 = strongly disagree, 2 = disagree, 3 = neither disagree nor agree, 4 = agree, 5 = strongly agree): (1) police traffic stops and searches help reduce crime, (2) police traffic stops and searches help take criminals off the street, (3) police traffic stops and searches help take drugs and firearms off the street, and (4) police traffic stops and searches are an effective crime control tactic (alpha = 0.923). Binary variables were used to denote which condition respondents received (1 = yes, 0 = not receiving that condition) with the control condition serving as the reference category. While we present the uncontrolled model as our main analysis, we also present a controlled model incorporating various variables such as socio-demographics to assess which controls are related to the dependent variable and to sharpen model fit (Armitage, 1996). We controlled for criminal justice system legitimacy by combining four variables that asked respondents whether the criminal justice system (a) protect people’s basic rights, (b) are generally honest, (c) generally do their jobs well, and (d) can be trusted to do what’s right for my community (alpha = 0.93). Procedural justice was measured using a combined score of five items asking the respondent whether those in the criminal justice system (a) treat everyone equally, (b) clearly explain the reasons for their actions, (c) treat people with dignity and respect, (d) treat people fairly, and (e) respect people’s rights (alpha = 0.94) (Pryce et al., 2017). Variables incorporated in the controlled model include dummy variables of race (Black, Other, and White [reference category]),Footnote 4sex (1 = male), six categories of age,Footnote 5political beliefs (1 = liberal, 2 = moderate, 3 = conservative), six categories of education attainment,Footnote 6married (1 = yes), employed (1 = yes), and crime victim (in past 12 months) (1 = yes). Sample characteristics and characteristics by condition can be found in Table 1 below and Table A1 in the Appendix, respectively. We also report theoretically relevant interaction terms to assess if various sub-groups were more likely to be impacted by the treatments.

Table 1 Sample characteristics
Full size table

Results

Uncontrolled and controlled models

Table 1 shows the sample characteristics. The average respondent was an employed, married, White female, who is politically moderate, with a 4-year college degree, and has not been a crime victim in the past year. Table 2 presents our uncontrolled OLS regression model. Coefficients are average marginal effects. In comparison to the control condition, respondents randomly assigned the contraband treatment were significantly (p = 0.004) more likely to support discretionary traffic stops as an effective crime-fighting tool (AME = 0.112, CI = 0.035, 0.189, p = 0.004). Although the coefficient was in a negative direction (AME =  − 0.025, CI =  − 0.102, 0.053), the disparity treatment had a non-significant (p = 0.532) effect on perceptions of traffic stops as a crime control strategy (compared to the control condition).

Table 2 Uncontrolled experimental results
Full size table

Findings from the controlled model (Table 3) confirm the above findings. That is, those assigned the contraband treatment were significantly more likely to support discretionary traffic stops as an effective crime-fighting tool (AME = 0.096, CI = 0.026, 0.166, p = 0.008). The effect of the disparity treatment was again non-significant (p = 0.150) in influencing respondents’ perceptions of discretionary traffic stops. Several control variables were significantly associated with the dependent variable. Those who held high views of the justice system as legitimate and procedurally just were significantly more likely to support discretionary traffic stops. Black respondents, males, those who identified as “other” in their gender, and those with greater educational attainment were significantly less likely to support discretionary traffic stops, while older respondents and those who lean politically moderate and conservative were more likely to support discretionary traffic stops.

Table 3 Controlled experimental results
Full size table

Multiplicative models assessing treatment heterogeneity by race of respondent

Since research has found that Black drivers are disproportionately pulled over by police (Baumgartner et al., 2017; Pierson et al., 2020), we assess whether our experimental treatments differentially impacted Black respondents (compared to White respondents). We are not making claims of causal moderation (Bansak, 2021); instead, we are testing for whether there was treatment heterogeneity by race of respondent. Point estimates and p-values in the presence of an interaction, even in linear models may be misleading (Busenbark et al., 2022); therefore, we use Stata’s margins command to estimate marginal effects and employ graphical methods to tease out the nature of the interactions (Bartus, 2005; Williams, 2012).

First, we examine differences between race by experimental condition on perceptions of discretionary traffic stops (Fig. 1 above and Table A2 in the Appendix). Across all experimental conditions, there were statistically significant differences between Black and White respondents’ perceptions of discretionary traffic stops. That is, regardless of experimental condition, Black respondents had significantly lower support of discretionary traffic stops compared to White respondents. Note that for the contraband treatment (as evidenced by Fig. 1), the difference in perceptions of discretionary traffic stops between Black and White respondents was smallestFootnote 7.

Fig. 1
figure 1

Racial differences by condition on perceptions of traffic stops

Full size image

Now we turn to differences within race that vary by experimental condition (Fig. 2 below and Table A3 in the Appendix). We find that there was a non-significant difference between White respondents who received the disparity treatment and White respondents who received the control condition in their support (or lack of) discretionary traffic stops (p = 0.511). Similarly, there were no significant (although marginally, p = 0.079) differences between Black respondents assigned the disparity treatment and Black respondents assigned the control condition in their support for discretionary traffic stops. While there was not a significant difference between White respondents assigned the contraband treatment compared to White respondents assigned to the control condition, there were significant differences between the Black respondents assigned the contraband treatment compared to Black respondents assigned to the control condition. That is, Black respondents who were randomly assigned the contraband treatment were significantly (p = 0.003) more likely (AME = 0.295) to support the use of discretionary traffic stops in controlling crime compared to Black respondents assigned the control condition. We discuss the implications of our findings below.

Fig. 2
figure 2

Within race differences by experimental condition on perceptions of traffic stops

Full size image

Discussion

Since traffic stops remain the most frequent iteration of police-civilian contacts in the United States (Bureau of Justice Statistics, 2018), some of which may result in a search, we sought to explore public perceptions of their use in combatting crime when presented with information on the outcomes of such stops. As such, we executed an original statewide survey experiment of South Carolinians to assess whether data on outcomes of discretionary traffic stops impacted public support for their utility in combatting/suppressing crime. Results from 4740 respondents revealed that in comparison to the control group, respondents assigned the contraband treatment were significantly more likely to support the utility of discretionary traffic stops in combatting crime. Although the coefficient was in a negative direction, there was a non-significant effect of the disparity treatment on respondent’s perceptions of discretionary traffic stop efficacy. Additionally, compared to White respondents, Black respondents assigned to the contraband treatment were significantly more likely to support discretionary traffic stops as a crime prevention strategy. We discuss our findings in further detail and outline policy implications below.

Respondents who received statistics on a 25% contraband hit rate were significantly more likely to support police use of discretionary traffic stops as an effective crime control strategy. That is, the public may lend support to discretionary traffic stops as an adequate tool to combat jurisdictional crime rates when one in four traffic stop searches results in a contraband seizure. As evidenced by our findings, the public may be swayed by the outcome data from traffic stop searches that remove dangerous contraband from their community. This finding aligns with the sentiment of some police, police administrators, and scholars who argue that discretionary traffic stops are a key crime-suppressing tool as they (1) increase presence in the community, (2) may result in contraband seizure, and (3) lead to arrest of potentially violent criminals (Lee, 2022; Wu & Lum, 2019). These findings also align with other studies that view police as playing a vital role in combatting crime (Boehme et al., 2023; Braga & Weisburd, 2020). While outside the scope of this study, these findings may lend support for other policing strategies/tactics that result in crime reductions (e.g., hot spots, focused deterrence), which, if presented to the public, may result in more positive perceptions of police and conducting crime control strategies. Increasing public support of police may have important implications for police officer morale. That is, as the second Peelian Principle suggests, for police to execute their daily duties, the public must approve of their actions. Additionally, the President’s Task Force on 21st Century Policing’s first pillar proposes that encouraging police legitimacy may impact the public’s willingness to obey the law. Demonstrating police effectiveness (in this case during discretionary traffic stops) may encourage public approval of law enforcement, which has a secondary benefit in enhancing police morale in executing daily policing duties.

We also found that the racial disparity treatment did not significantly impact respondents’ perceptions of discretionary traffic stops. This finding is particularly interesting in a time when police are experiencing a “legitimacy crisis” (Todak, 2017) and calls to abolish/defund the police are emerging, mostly due to suggestions of racially disparate policing. Note that the statistics were arguably “stronger” (e.g., Blacks were 55% more likely to be pulled over than Whites, compared to the 25% hit rate in the contraband condition) in the racial disparate condition, with the added element that contraband was less likely to be found on Black drivers (in comparison to White drivers). A couple of conclusions may result from this non-significant finding. Although we presented racial disparity statistics, respondents may have believed that the potential crime deterrent/suppressant effect of discretionary traffic stops (as asked in the measures of the dependent variable) may have still outweighed the effect of racial disparities resulting from discretionary traffic stops. Much empirical evidence and societal knowledge exists regarding disparities in police traffic stops. The public may have already been “aware” of such disparities, whereby disparity information may have already been “normalized” so as to not sway respondent priors regarding the efficacy of such types of stops. That is, respondents may have weighed a “cost” of disparate stops and a crime control “benefit” that may result from such stops.

When interacting the contraband condition by race, we found that Black respondents were significantly less likely to support discretionary traffic stops as a crime control tactic, regardless of which condition they were assigned. This finding is not surprising, as it is well-documented that Black Americans compared to White Americans are more likely to feel targeted during traffic enforcement (Epp et al., 2014). Note that throughout all conditions, Black respondents held lower support for discretionary traffic stops compared to White respondents. With that said, we also found that Black respondents assigned the contraband treatment (compared to Black respondents assigned the control condition) were significantly more likely to support traffic stops as a crime prevention tool. It is well-documented that Black Americans are disproportionately more likely to experience violence either vicariously or directly (Fowler et al., 2015); therefore, they may be impacted by data that shows the potential crime prevention outcomes of certain police strategies (e.g., discretionary traffic stops) that may make their communities safer. Said differently, the outcomes of stops may outweigh the process and mechanisms that may occur during interactions with police during stops. Additionally, Black respondents may be particularly sensitive to data on crime prevention since these policing tactics are not just theoretical but have tangible, everyday impacts on their safety, freedom, and well-being.

While research highlights that Black Americans feel over-policed in general (Brunson, 2007), and specifically during traffic stops, the concern of under-policing and neglect also concerns the Black community (Boehme et al., 2022). Therefore, respondents assigned to the contraband treatment may internalize the contraband outcome data as police being present, serving, and addressing crime problems within the community. Thus, information of discretionary traffic stop hit rates may then signal to Black respondents that the community is not neglected and left vulnerable and that the outcomes of stops may show police are addressing the crime problem. These findings also align with some of the literature on Black American’s attitudes towards crime control policies and practices (Metcalfe & Pickett, 2018; Wilson & Dunham, 2001). Black American’s perceptions of crime control policies and policing are undoubtedly complex. For example, prior research has found that Black Americans may support crime control policies but also yearn for fair and equitable criminal justice treatment (Ramirez, 2015; Rios, 2011). In the case of the present study’s goal, which was to assess public support of discretionary traffic stops, we found evidence that Black respondents assigned the contraband treatment were more likely to support this specific crime control tool. However, if we asked about perceptions of equitable and fair discretionary traffic stops, we may have found nuance in both support of discretionary traffic stops as an effective crime tool but also varying perceptions of fairness. However, we leave this line of investigation for future researchers.

As for policy implications, the broad exercise of police discretion, particularly during traffic stops, remains a fundamental aspect of police work. In evaluating the effectiveness of these policing strategies, it remains crucial for agencies to consider their differential impact on various communities, especially those historically more vulnerable to violence and exposure to police presence. Importantly, such evaluations must acknowledge and address the complexities highlighted by studies like ours. Our findings indicate that, despite the presence of disparities, local communities may still support practices like discretionary traffic stops as they may view them as vital in controlling crime and addressing their own vulnerability to criminal victimization. Striking a balance between these nuanced perspectives is crucial in developing policing policies that not only consider community safety concerns but also firmly reinforce the need to reduce disparities in police practices.

Police administrators and scholars must consider ways in which these stops can continue without eroding community trust in police. Police must consider ways to increase “hit rates,” present data/statistics to the public in a digestible way to improve legitimacy and be transparent with the public in discussing the value of certain strategies. While we found that the contraband treatment had a significant effect on Black respondents, we caution policymakers and practitioners in engaging in simply executing more discretionary traffic stops, since there is a nuanced relationship between crime control policies and perceptions of fairness of such policies among Black Americans. Instead, we suggest strategic evidence-based practices in high-crime areas may result in greater community satisfaction, with some evidence of enhanced support in Black communities (Metcalfe & Pickett, 2018).

Research shows that Americans do not support the abolish/defund police movement including people of color (Vaughn et al., 2022). Based on our findings, we suggest that police pulling back on discretionary stops and to an extent de-policing may not be supported by the public, particularly if it has implications for community safety. Therefore, we suggest that police administrators are transparent with the community in their crime prevention efforts, with emphasis that officers are present and engaging in crime prevention efforts, highlighting any positive effectiveness of such efforts in reducing jurisdictional crime rates. While it is obvious that police can improve their relationship with the community through procedural justice and legitimacy (Sunshine & Tyler, 2003), transparency with the public regarding community safety efforts and outcomes of strategies may help build a positive relationship with the community.

Although our study design allows for causal conclusions due to the experimental design, there are some limitations worth noting. While we are confident in the internal validity of our results, we are cautious in extrapolating our findings nationally and within certain populations of the state of South Carolina. That is, our sample was mostly employed, White, female, who are employed; therefore, potentially impacting our results. Historically, South Carolina is a politically conservative state, and therefore, our findings may not apply to other areas of the country; however, this study may shed light on public perceptions towards traffic stops of those from historically conservative states. Finally, while we present findings of search hit rates and racial disparities derived from one study, the public may also be swayed by other data from other studies on different outcomes of traffic stops (e.g., effect on traffic collisions and fatalities). Although the contraband treatment is derived from an actual study that found a 25% hit rate, some studies do not always find a hit rate as high as 25% (Engel & Calnon, 2004). Therefore, respondents in our sample may not have been as strongly swayed had our contraband treatment stated a lower hit rate. Additionally, providing different stop outcomes (e.g., use of force) may have differentially impacted our findings. Future research should uncover other factors that may affect public approval (or disapproval) of commonly implemented police tactics. As stated above, we only asked about support for discretionary traffic stops as an effective crime control tool, not whether these stops are fair and equitable. Future research should close the gaps in this research. Therefore, practitioners should be cautious when utilizing findings from this study as a clearcut indicator that Black Americans completely support discretionary traffic stops.

Following the Summer of 2020, reform activists and politicians proposed several reform efforts to restrict police discretionary traffic stops. While there has been much debate among activists and practitioners over the consequences of policies that restrict such stops, little is known regarding public support of discretionary traffic stops and whether presenting data on stop outcomes influences public support for stops. We have no reason to think police traffic stops will vanish in the near future; therefore, understanding ways in which the public may support such a contentious policing tactic is worth exploring. We found that when the public is presented with data on contraband hit rates, an outcome that may contribute to community safety, they are more likely to support such a tactic. We suggest police sharing data on outcomes of police strategies with the public, particularly of commonly implemented strategies, to engage and potentially improve public support for such policies. This study contributes to the literature on public attitudes towards discretionary traffic stops, their utility in enhancing community safety, and ways police data may shape public perceptions.