The third was an officer struggling with panic symptoms after responding to a child-abuse case.
Malcolm treated each exactly as he would have before the park.
That required effort.
The encounter with Prescott was still circulating in his body.
He could feel it in his shoulders.
At 5:30, he called his wife from the parking garage.
She answered on the second ring.
“You sound tired.”
“I had an interesting lunch.”
“What happened?”
Malcolm told her.
Not dramatically.
He simply described it.
When he finished, she was silent for a long time.
Then:
“Are you okay?”
“Yes.”
“That was too fast.”
He sighed.
“I’m angry.”
“That sounds more honest.”
“He verified me only after I told him I worked with police.”
His wife understood immediately.
“And if you didn’t?”
Malcolm looked through the windshield.
“That’s the part I can’t stop thinking about.”
Internal Affairs began with the easiest evidence.
The body camera.
The dispatch audio.
The computer-aided dispatch log.
Investigators did not begin by deciding Prescott had racially profiled Malcolm.
They began by asking what could be proven.
At 12:17:04, Prescott’s body camera showed him entering the park.
At 12:17:19, Malcolm became visible in the frame.
At 12:17:28, Prescott altered direction toward the bench.
At 12:17:41, the first demand for identification occurred.
No dispatch call directed Prescott toward Malcolm.
No active incident involved a suspect matching him.
No park employee had requested assistance.
No citizen complaint identified a suspicious Black man.
Prescott had initiated the contact independently.
That alone was not misconduct.
Police officers were allowed to speak with people.
The question was what happened next.
Investigators watched the footage again.
Then again.
They transcribed every exchange.
Prescott repeatedly invoked “reports” and “thefts,” but he did not identify any specific incident connected to Malcolm.
He told Malcolm to remain where he was.
Then, when Malcolm asked whether he was detained, Prescott avoided a clear answer before eventually saying Malcolm was not free to leave until his identity was established.
That was the point the legal review focused on.
A department attorney wrote in the preliminary memo:
The initial contact appears consensual. The encounter may have converted into a detention when Officer Prescott instructed Pierce that he was not free to leave pending identification. The presently available record does not identify specific articulable facts connecting Pierce to criminal activity at the time that instruction was given.
That language was careful.
No slogans.
No outrage.
Just the record.
Then investigators listened to dispatch.
“Send me another car. Suspicious male.”
Dispatch asked:
“What’s the nature?”
“Subject refusing identification during investigation.”
“Do you have a related call number?”
“Negative. Self-initiated.”
That exchange became central.
The next step was not Malcolm’s psychological opinion.
It was Prescott’s data.
Internal Affairs pulled five years of stop reports.
The raw numbers immediately looked unusual.
1,247 documented stops.
892 involving people recorded as Black, Latino, Asian, Middle Eastern, or another non-white category.
355 involving white subjects.
That was roughly 71.5% versus 28.5%.
But investigators knew raw percentages could mislead.
Prescott did not patrol the entire city evenly.
His assignments mattered.
Crime patterns mattered.
Shift times mattered.
Geography mattered.
A patrol officer working predominantly in a diverse district would naturally have contact demographics different from citywide population numbers.
So Internal Affairs sent the data to the department’s crime-analysis unit and an outside statistician already under contract with the civilian oversight board.