Night and Day: Racial Disparities in Arrests across Black and Latine Communities in 2020 through the Veil of Darkness

Background: There is a link between a driver’s race and gender, and traffic stops in the U.S. with male Black and Latino drivers being arrested at higher rates than white drivers. Additionally, time of day of a traffic stop has an effect on whether a Black or Latino driver is arrested and the current study further investigates this with the veil of darkness hypothesis which posits that a driver’s race is less likely to be identified by police before a traffic stop during nighttime hours compared to daytime hours, and assumes that this change is due to racial profiling.

Methods: The sample (n=44,506) was from the 2020 Racial Profiling Dataset accessed through the City of Austin, Texas open data portal. Data was analyzed using a binomial logistic regression with arrest vs. no arrest as the dependent variable. Covariates included sex, street type, time of stop, and the reason for a traffic stop. An interaction with time of day and race/ethnicity was added to further examine the veil of darkness hypothesis. Finally, predicted probabilities were calculated for risk of arrest.

Results: Results of binomial logistic regression showed that Black drivers (OR=3.17, 95% C.I. [2.71, 3.69]) and Latine drivers (OR=2.28, 95% C.I. [1.99, 2.62]) had higher odds of being arrested compared to white drivers and that male drivers being more likely to be arrested (OR=1.56, 95% C.I. [1.44, 1.70]) compared to female drivers. Drivers stopped for violations other than moving traffic laws had a significantly higher chance of being arrested (OR=5.79, 95% C.I. [5.36, 6.25]) compared to those who were stopped for moving traffic violations. Finally, being stopped on a city street was associated with a higher likelihood of arrest (OR=1.62, 95% C.I. [1.48, 1.77]) compared to being stopped on a highway The interaction between nighttime and race/ethnicity was significant, during night stops, Black drivers (OR=0.38, 95% C.I. [0.31, 0.47]) and Latine drivers (OR=0.80, 95% C.I. [0.67, 0.95]) were less likely to be arrested compared to white drivers. For all racial/ethnic groups, the probability of arrest is higher at night than during the day, but predicted probabilities show that Latine drivers are more likely to be arrested at night compared to all other racial/ethnic groups.

Conclusions: Racial profiling during traffic stops must not be seen as an individual issue but as a public health issue, especially in light of study findings validating the veil of darkness hypothesis. The impacts of racialized policies must be examined to see whether these increase racial profiling by police for Black and Latine communities. Data must be collected when police ask about documentation status and the resulting outcomes of this documentation investigation. Social workers should ensure that Black and Latine community voices are equitably integrated into research and policy creation through approaches like community-based participatory research, town halls, and policy community advisory boards. Anti-racist approaches and policies to reduce police/public interactions should be implemented to reduce outcomes that disproportionately impact Black and Latine communities.