California passed the RIPA in 2015 because of concerns about bias in law enforcement that lead different groups to have different law enforcement experiences. Beyond the obligation to collect data on controls, legislation has broadened and clarified the definition of racial and identity profiling to take into account and build on protected group status, such as race and gender. ethnicity, to “… decide who to screen or to decide on the scope or substance of law enforcement activities following a judgment …”
Research consistently finds evidence of racial bias, explicit and / or implicit, widely in society (see e.g. Bertrand and Mullainathan 2004; Bayer et al. 2017; Rothstein 2017; Avenancio-Leon and Howard 2019; Chetty et al. 2020 ; Kline, Rose and Walters 2021). In addition, research has also revealed racial discrimination within the criminal justice system in the decisions of juries, judges and prosecutors (Anwar, Bayer and Hjalmarsson 2012; Arnold, Dobbie, Yang 2018; Sloan 2019). It is perhaps not surprising that these racial prejudices extend to policing (Fryer 2019; Luh 2019; Hoekstra and Sloan 2020; Ba et al. 2021; Feigenberg and Miller 2021; Goncalves and Mello 2021) , providing support to concerns historically raised by communities of color. —Resolved concerns following the murder of George Floyd.
Many factors contribute to whether an agent arrests someone and the agent’s subsequent actions. And while the RIPA data points quite strongly to differences in judgment results by race and ethnicity (RIPA Board Report 2021), these differences may echo circumstances that do not reflect a individual agent. The reason and context of the stop likely influences an officer’s decisions and actions – for example, an officer may simply warn a driver stopped for speeding. Therefore, differences in quitting experiences between blacks and whites may reflect differences in the reasons for quitting (Figure 2).
Regardless of race and ethnicity, if an officer has observed a person committing a crime, if a person has a warrant or a weapon, that person will likely be detained and searched, and possibly incarcerated after a stop. Such situations can be more confrontational – including the potential for the use of force – than a traffic stop. If a person acts erratically, perhaps due to behavioral health issues, an agent can change their decisions and actions. The prevalence of such situations by race and ethnicity may contribute to differences in outcomes.
Additionally, younger / inexperienced drivers may be more likely to break the rules of the road and are therefore more likely to be stopped than older, more experienced drivers. With men and adolescents / young adults involved in relatively more criminal activity, officers may monitor young men more when they are arrested than older people or women, regardless of race / ethnicity ( for example, Ulmer and Steffensmeier 2014).
On average, black Californians arrested by law enforcement are perceived to be younger than white Californians who are arrested (Technical Annex, Table A2). The people arrested are also more often men. Relatively higher proportions of black people are arrested on reasonable suspicion, a pending arrest warrant, or mandatory supervision of a parolee or probationer. The latter two categories make a search more likely because a search warrant is not, nor is, necessary in California. The officers also report having visibly seen contraband in a higher proportion of arrests of blacks than of whites.
Among the 15 largest law enforcement agencies, the California Highway Patrol made more than 60% of all stops in 2019 of white individuals, but only about 35% of stops of black individuals. And while the Los Angeles Police Department (LAPD) accounts for nearly 31% of checks on black Californians, the agency only performed 10% of checks on white Californians; partly reflecting that a higher share of Los Angeles’ population is black (about 8%) compared to the state as a whole (about 6%). Agency-level differences in policing strategies, missions and roles, as well as officer behavior and biases, are also possible contributing factors.
Recognize differences in personal traits and contexts
Differences in contexts, locations and agencies likely contribute to racial disparities in judgment outcomes. Our goal here is to use regression models that adjust to account for racial / ethnic differences in such factors, and steer us toward more “apple-to-apple” comparisons.
That is, we seek to compare the results of stops by race / ethnicity for, say, individuals of the same age and sex, arrested for the same reasons by a law enforcement agency. the law given. We also adjust whether the officer reported seeing contraband and whether the person had a pending warrant or is on parole or probation. (See Technical Annex B for a detailed discussion of the analysis and regression model.)