They controlled for patrol area.
Shift.
Call type.
Self-initiated versus dispatched contact.
Neighborhood demographics.
Reported suspect descriptions.
Outcome of the contact.
The adjusted analysis took three weeks.
The pattern did not disappear.
It became more specific.
Prescott’s dispatched contacts roughly resembled those of comparable officers working the same beats.
His self-initiated stops did not.
When Prescott chose whom to investigate without a dispatch call, Black men were significantly overrepresented compared with similar officers working the same locations and hours.
The disparity was greatest in Riverside Park and two nearby commercial corridors.
Then analysts examined outcomes.
Stops involving white subjects were more likely to begin with a documented violation, witness description, or observed infraction.
Stops involving Black subjects were substantially more likely to begin with broad labels such as suspicious behavior, loitering, casing area, furtive movement, or officer observation.
Those words were not automatically illegitimate.
Sometimes they described real circumstances.
But frequency mattered.
Context mattered.
Patterns mattered.
The investigators then reopened Prescott’s two previous citizen complaints.
The delivery driver had been Black.
The college student had been Latino.
Both encounters had been self-initiated.
Both complaint investigations had lasted less than forty-eight hours.
Neither investigator had reviewed Prescott’s broader stop history.
One disposition read:
Officer demeanor may have been perceived as abrupt but remained within policy.
Another said:
Insufficient evidence to establish discriminatory motivation.
Technically, both findings could have been defensible at the time.
The problem was that no one had ever put the files beside one another.
Internal Affairs Lieutenant Rachel Monroe wrote in her case notes:
The concern is not that earlier investigators necessarily reached unreasonable conclusions on isolated complaints. The concern is that the department possessed multiple data points and lacked a mechanism requiring them to be evaluated collectively.
That sentence would later appear in the civilian oversight report.
Prescott was notified of the investigation on the third day.
His union representative attended the interview.
He sat at a conference table in a gray department polo instead of a uniform.
The body camera footage had already been preserved.
“Why did you approach Dr. Pierce?” the investigator asked.
Prescott leaned back.
“He looked out of place.”
The investigator waited.
Prescott realized the phrase sounded bad.
“I mean his behavior caught my attention.”
“What behavior?”
“He was sitting alone watching the area.”
“He was reading.”
“He had a book.”
“Did you observe him watching parked vehicles?”
“I believed he may have been.”
“Did your body camera capture that?”
“Not before I approached.”
“Was there a recent theft call in Riverside Park?”
“There had been crime in the district.”
“That wasn’t my question.”
Prescott shifted.
“No active theft call.”
“Was there a suspicious-person call?”
“No.”
“Did anyone give you Dr. Pierce’s description?”
“No.”
“Then why did you tell him there had been reports of suspicious activity?”
“Because there had been.”
“Where?”
“In the general area.”
“What report specifically were you investigating?”
Prescott looked at his union representative.
The representative remained silent.
“I was conducting proactive patrol.”
The investigator nodded.
“That’s a different answer.”
Prescott’s expression hardened.
“I don’t see how.”
“One implies you were responding to reported activity. The other means you selected Dr. Pierce based on your own observations.”
“I’m allowed to do that.”
“Yes.”
The investigator leaned forward.
“We’re trying to determine what you observed.”