Police departments cannot afford to ignore their own RIPA data
Sacramento’s 2024 stop data demonstrate why agencies must examine disparities, validate their conclusions and identify problems before someone else does it for them
California law enforcement agencies devote substantial time to collecting stop data required by the Racial and Identity Profiling Act. Officers enter information about the perceived characteristics of the person stopped, the reason for the stop, actions taken, searches, evidence discovered, property seized and the final disposition.
But collecting the data is only the first obligation.
Police departments also need to understand what their data reveal—and what they do not. An agency that waits until an advocacy organization, journalist, attorney or oversight body publishes an analysis is already operating from behind. Its leaders may be forced to respond publicly before they know whether the published statistics are accurate, whether the correct population was analyzed or whether their own data contain legitimate warning signs.
That is the broader lesson from my recent report, Beyond Disparity: A Practical Review of Sacramento PD RIPA Stop Data. The report evaluates the Sacramento Police Department’s 2024 RIPA records in response to the ACLU of Northern California’s May 2026 report, Driving While Black and Brown.
The purpose of my review was not to make Sacramento’s disparities disappear. They do not. Nor was it to presume that every disparity established discriminatory enforcement. It does not.
The purpose was to test the claims against the data, use more precise analytical populations and identify where statistical findings should lead to a case-level audit.
Disparity is a starting point, not a verdict
The ACLU reported that Black people constituted approximately 12% of Sacramento’s population but 33% of its traffic stops. It also reported higher search rates for Black and Latino drivers and concluded that the disparities painted a clear picture of racially biased, pretextual enforcement.
Those findings raise legitimate questions. The dispute is not whether a disparity exists. The more difficult question is what the disparity proves.
Residential population is a useful screening benchmark, but it is an imperfect measure of traffic-enforcement exposure. It does not account for who is driving, where and when people travel, nonresident motorists, vehicle condition, actual violation rates, calls for service or the geographic deployment of officers.
A population comparison can identify a difference requiring investigation. By itself, however, it does not tell us why the difference occurred.
That distinction is important. Police leaders damage their credibility when they dismiss a disparity merely because the data do not prove discrimination. Advocacy groups and commentators make the opposite error when they treat an aggregate disparity as conclusive proof of officer motivation or an equal-protection violation.
Sound analysis requires more discipline from both sides.
Know what is being counted
My Sacramento review began with a basic question: What does each row of RIPA data represent?
The 2024 dataset contained 24,104 person records in which a traffic violation was coded as the reason for the stop. That figure replicated the ACLU’s reported 2024 traffic-stop count. But those records represented 23,126 distinct stop identifiers—not 24,104 unique driver stops.
RIPA is primarily person-level data. A vehicle stop involving a driver and a passenger may produce more than one reportable person record. Treating every row as a separate driver stop can make the language of an analysis more precise than the underlying data.
I therefore developed a narrower population consisting of traffic-violation, vehicular, officer-initiated, non-passenger, first-person records. That population contained 21,208 apparent-driver stops.
The refined denominator improved analytical accuracy, but it did not explain away the disparity. Black individuals constituted 31.2% of the apparent-driver population, compared with 31.9% in the broader traffic-violation population.
That is an important finding in both directions. The original denominator was broader than unique driver stops but correcting it did not materially change the racial distribution. A responsible review must disclose both facts.
Search disparities need more than one statistic
Searches present an even greater analytical challenge.
Among Sacramento’s apparent-driver stops, the full-search rates were:
10.8% for Black drivers;
7.4% for Hispanic or Latino drivers;
4.6% for White drivers; and
3.9% for Asian drivers.
I also conducted an adjusted analysis controlling for measured age, grouped gender, traffic-violation type, nighttime and weekend stops. Black apparent drivers had approximately twice the odds of receiving a full search as White drivers. Hispanic or Latino drivers had approximately 24% higher odds.
The Black-White difference was statistically significant and operationally meaningful. It is a legitimate audit signal that should not be minimized.
It is not, however, a complete explanation of why the searches occurred.
Parole, probation or another supervision-related authority was recorded in 59% of full searches of Black apparent drivers, 54% of White-driver searches and 46% of Hispanic or Latino-driver searches. Removing searches carrying that basis substantially reduced the absolute search rates, although the relative Black-White disparity remained.
Supervision status therefore provides important context, but not a complete answer.
The next step should be a review of individual cases. When did the officer learn of the person’s supervision status? Was the search condition confirmed? Did the officer know about it before initiating the stop or only afterward? Was the initial enforcement action independently supported? What do the report, citation, dispatch history and body-worn video show?
RIPA data can direct an agency toward those questions. In most cases, they cannot answer them alone.
Search likelihood and search productivity are different questions
The ACLU reported that 90% of searches resulted in no property or contraband being seized. That figure presents another lesson for police analysts: Definitions and denominators matter.
Discovery and seizure are different RIPA outcomes. An officer may discover contraband or evidence without the event being coded as a seizure in the way a particular analysis defines that term. Productivity rates must also be calculated among searched persons, not among all stopped persons.
In Sacramento’s apparent-driver population, 30.2% of full searches recorded a discovery, while 15.8% recorded a seizure. After adjustment for measured factors, race and ethnicity were not statistically significant predictors of discovery.
This does not establish that officers applied equal thresholds when deciding whom to search. A department cannot use a discovery rate to erase a disparity in search decisions.
It does show that search likelihood, discovery and seizure are separate questions. Each tells command staff something different:
How frequently was each group searched?
What legal authority supported the search?
How often did the search produce contraband or evidence?
What was seized?
Did productivity vary by officer, unit, violation, location or search basis?
Collapsing these questions into a single statistic may generate an effective headline, but it does not produce a complete operational assessment.
Not every result favors the agency
An internal RIPA review should not begin with a predetermined goal of rebutting criticism. If the purpose is merely to defend the department, analysts will lose credibility and may overlook real problems.
My Sacramento analysis identified several areas warranting targeted review. Window-obstruction enforcement stood out.
The apparent-driver dataset contained 1,090 window-obstruction stops. Black and Hispanic or Latino drivers constituted approximately 35% and 40%, respectively, of those stops. Black drivers had a 16.5% full-search rate, compared with 8% for White drivers.
Those numbers do not establish that window-obstruction laws were used as a racial pretext. They do justify examining the cases more closely, particularly stops followed by a search, prolonged detention, force or no citation, warning or arrest.
A useful audit would compare similarly situated stops within the same Vehicle Code section, locations, time bands, officers and enforcement units. It should evaluate whether the violation was documented consistently, why a search followed and whether enforcement outcomes complied with policy and supervisory expectations.
A department willing to perform that examination is not conceding unlawful conduct. It is demonstrating that its conclusions are based on evidence rather than institutional instinct.
Build a recurring RIPA review process
RIPA analysis should be an ongoing management function, not an emergency response to public criticism.
At a minimum, agencies should establish a quarterly or semiannual review that follows the enforcement decision pipeline:
Stop population → reason for stop → violation → actions taken → search authority → discovery or seizure → citation, warning, arrest or no action
The review should distinguish among person records, distinct stops, drivers, passengers, calls for service and officer-initiated activity. Analysts should examine raw disparities and, when the data permit, comparisons accounting for age, gender, violation type, time, location, officer, unit and legal search authority.
Automated quality checks should identify:
Multiple person records associated with one stop;
Missing search bases;
Searches recorded without a corresponding action;
Impossible or inconsistent field combinations;
Unusual stop durations;
Abrupt changes in reporting practices;
Officers or units with patterns materially different from comparable assignments; and
Enforcement categories combining high discretion, substantial disparity and low yield.
Agencies should also be cautious about comparing themselves with another department. In my report, Sacramento and Oakland had the same 30.2% discovery rate among apparent-driver full searches and the same 4.4% arrest rate. Yet their traffic-enforcement mixes and outcome coding differed substantially.
Peer comparisons may help identify unusual patterns, but differences in policy, deployment, enforcement priorities and coding can make another agency a poor control group.
Use RIPA data to guide body-worn camera audits
One of the greatest benefits of routine RIPA analysis is that it helps departments identify—and narrowly frame—questions the data cannot answer by themselves.
RIPA data can show that one group was searched more frequently, received a particular enforcement action more often or experienced a lower rate of citations or contraband discoveries. It generally cannot establish the complete legal and factual circumstances of an individual stop. It does not show everything the officer observed, the sequence in which information became known, how the officer communicated with the driver or whether similarly situated motorists were treated consistently.
Because most agencies now use body-worn cameras, departments have a practical means of moving from statistical patterns to case-level review. The data can be used to select a random or stratified sample of stops from a narrowly defined category and then compare the RIPA record with the body-worn camera recording, incident report, citation and dispatch history.
For example, an agency could randomly sample traffic stops of Black drivers for window obstruction and examine:
Whether the recorded obstruction was visible and supported the stop;
How the officer described the reason for the stop;
Whether additional enforcement or investigative activity followed;
What facts supported any detention, frisk or search;
Whether consent was requested and voluntarily given;
Whether the RIPA record accurately reflected the encounter; and
Whether comparable stops were handled consistently across officers, units and demographic groups.
A properly designed audit should not presume misconduct. It should test whether the statistical signal corresponds with lawful, consistently applied enforcement—or whether it reveals a pattern requiring additional review.
When a department identifies discrepancies through its own audit, openly acknowledges them and takes corrective action, its transparency carries greater credibility than a response issued only after outside criticism. Repeated findings may expose unclear policy, inconsistent supervision, deficient data entry or a need for focused training. The review may also confirm that officers acted lawfully and professionally under circumstances that the RIPA fields could not capture.
Regular auditing creates accountability in both directions. Officers who know that selected encounters may be compared with their RIPA submissions and body-worn camera recordings have an additional reason to remain within policy and document their actions accurately. Officers who make mistakes can be corrected before a pattern develops. Those whose recordings demonstrate sound judgment, professionalism and accurate reporting can be recognized for good work.
RIPA analysis is therefore most useful when it does more than produce annual percentages. It should help a department decide which encounters to examine, which questions to ask and where supervision or training can improve performance.
Data literacy is part of organizational readiness
Police executives do not need to become statisticians, but they should be able to ask informed questions about their own data.
What is the denominator? Are we counting people, stops or drivers? Is the comparison descriptive or adjusted? What information is missing? Does the result identify a disparity, a low-yield practice or possible differential treatment? What additional records must be examined before reaching a conclusion?
Agencies should be prepared to defend sound practices when public claims go beyond the evidence. They must be equally prepared to correct practices when their own analysis identifies an unjustified disparity, an inconsistent policy application or a data-quality failure.
Those responsibilities are not in conflict.
The proper command response to RIPA data is neither automatic denial nor automatic condemnation. It is disciplined analysis, transparent acknowledgment of limitations and targeted review of the places where discretion and disparity intersect.
If police departments do not develop that capacity themselves, someone else will interpret the data for them—and the agency may not understand the evidence well enough to know which criticisms to challenge and which ones demand action.
References
American Civil Liberties Union of Northern California. Driving While Black and Brown: The Case for Banning Racially Biased Traffic Stops in Sacramento. May 2026. https://www.aclunorcal.org/app/uploads/2026/05/FINAL-Sacramento-Pretext-Stop-Report_05.26.pdf
Vernon, Paul B. Beyond Disparity: A Practical Review of Sacramento PD RIPA Stop Data—Descriptive Findings, Comparative Context, Limits of Inference, and Audit Recommendations. West Coast Police Integrity Advisors LLC, July 2026.
Author biography
Paul B. Vernon is a law-enforcement analyst and the principal of West Coast Police Integrity Advisors LLC. His work includes RIPA data analysis, police-practices review, statistical analysis and consultation concerning law-enforcement policy and accountability. He previously served in law enforcement and provides independent analytical and expert-witness services to public agencies and attorneys.