Law enforcement doesn’t always pull you over for the reason they say they do. A pretextual traffic stop is a policing tactic where an officer stops a vehicle for a minor traffic violation—which is the "pretext," or stated reason for the stop—with the underlying intention of investigating an unrelated crime. Drivers in pretext traffic stops are told they were pulled over for faulty brakelights, expired registration tags, or other low-level violations; once stopped, officers use their discretion to investigate drivers for unrelated crimes. As with other policing tactics, pretext traffic stops in San Francisco disproportionately target drivers of color.
Figure 1: Graph showing share of pretext traffic stops and share of population by race.
The overwhelming number of violations in pretext traffic stops are related to registration tags and rear license plates, with officers citing that they weren’t visible—an inherently subjective pretext.
Figure 2: Graph showing most common vehicle code infractions cited during pretext stops.
In recent years, serious efforts have been made to curtail this biased policing tactic. In 2023, the California legislature passed Assembly Bill 256, which prohibited all law enforcement officers from using expired registration tags as the sole basis of a stop, with limited exceptions. The state’s policy went into effect on July 1, 2024. San Francisco sought to be more ambitious and enacted DGO 9.07, which prohibited SFPD from stopping drivers solely based on pretext violations such as a failure to display or properly mount license plates and registration tags, rear taillight malfunctions, and more. The city’s policy went into effect on July 17, 2024
With over a year of data available since their passage, this research brief utilizes an Interrupted Time Series (ITS) regression model to analyze the effect of these two policies. To avoid the structural break caused by the COVID-19 pandemic, the analysis period was limited to stops from January 2022 through September 2025. This provided 132 weeks of data in the pre-intervention period and 63 weeks in the post-intervention period. The plot below shows general traffic violations trends—which include all traffic stops, not just pretext violations—across this timeframe, grouped into weekly totals.
Figure 3: Plot showing total weekly traffic stops.
Stops were then filtered to only include violations listed in AB 256 and DGO 9.07. To ensure that other factors, including fluctuations in police staffing, crime, and driving activity, weren't skewing the results, the analysis accounted for these variables when estimating the policies' impact.
The findings were stark—pretext stop bans produced immediate and substantial reductions in stops for Black, Hispanic, and White drivers, while stops for Asian drivers remained unchanged from their low pre-intervention baseline. Black drivers experienced a 47% decrease (from a counterfactual of 15.6 stops in the first week post-ban to 8.2 stops per week), Hispanic drivers saw a 68% decrease (7.4 to 2.4 stops per week), and White drivers experienced a 52% decrease (9.5 to 4.5 stops per week). Results were highly statistically significant across all racial groups except for Asian drivers.
Table 1: Table showing ITS Regression results.
The ITS model underscores that prior to the policy intervention, the number of weekly stops was increasing across all racial groups. There are, however, differing post-intervention trends. While Black and White drivers experienced sustained declines in stops, Hispanic drivers showed no significant continued trend after the initial decrease, suggesting the policies’ strong immediate effects were concentrated in the initial implementation period but weren’t sustained over time.
Figure 4: Plots showing weekly pretext stops and ITS results by race.
Further analysis was done to assess whether these large reductions in stops affected drug, gun, and contraband discoveries, as well as searches and arrests. Findings show the pretext stop bans resulted in negligible reductions to drug, gun, and contraband discoveries, while arrest rates stayed nearly the same. The only outcome that dropped with statistical significance was the number of searches conducted on those who were stopped for pretext violations.
Table 2: Table showing impact of pretext stop bans on discoveries, searches, and arrests.
When looking at these same outcomes by racial group we can see that for Black and Hispanic drivers, there were statistically significant reductions in searches. Put plainly, the pretext ban policies did not result in fewer guns and drugs discovered, nor did they result in the apprehension of fewer criminals. They did, however, result in fewer searches of Black and Hispanic drivers.
Table 3: Table showing impact of pretext stop bans on discoveries, searches, and arrests by race.
While it is clear there were significant immediate reductions in the number of stops for Black and Hispanic drivers, more research was conducted to determine whether the bans effectively removed racial disparities. Here, there were more nuanced results. While all pretext citations saw large reductions in their issuance, many continued to disproportionately be given to Black and Hispanic drivers.
Figure 5: Graph showing percentage of citations issued to Black and Hispanic drivers post-ban.
Finally, the data shows the possibility of SFPD substituting towards similar—but not explicitly banned—pretext stop violations after the bans took place. For example, citations for 26708.5(A),which prohibit window tints that alter or reduce light transmittance,were only given 12 times in the pre-ban period. Yet in the post-ban period, these citations were given out 47 times, a 291% increase.
Figure 6: Two graphs showing citations similar to those in pretext stop bans and the percentage of these citations issued to Black and Hispanic drivers.
It’s clear that the bans on pretext traffic stops are having their intended effect through dramatic reductions in the number of stops—and subsequent searches—of Black and Hispanic drivers in San Francisco. These reductions have not changed the rate of drug, gun, and contraband discoveries, nor have they changed arrest rates. Disparities continue to exist in the limited issuance of these pretext stop citations, however, and it is very possible that SFPD is substituting towards alternative pretext violations as a means of investigating drivers they are suspicious of. The city should consider other violations that fall under the scope of pretext stops, as well as the racial disparities in their issuance, to determine whether an expansion of DGO 9.07 would be beneficial.
The state of California should expand legislation that reduces the number of violations that law enforcement can use as the sole basis for a stop. Evidence from San Francisco strongly demonstrates the benefits of these policies, which dramatically reduce the number of stops and searches for drivers of color without hindering law enforcement’s ability to keep guns, drugs, contraband, and criminals off the streets. A statewide policy would go a long way towards removing bias in law enforcement and ending the over-policing of communities of color, all without compromising public safety.