It showed no attempt to strike anyone, reach for a weapon, or flee.
The video also captured Caldwell pushing Marcus's face against the desk.
The footage had no sound, but the dispatch recording supplied the missing words.
Investigators reviewed the radio traffic, access logs, personnel records, and the department's credential-verification procedures.
The findings were straightforward.
The department had an established process for verifying an employee's identity. The procedure required officers to contact the administrative command desk or dispatch when a person's credentials were unfamiliar or disputed.
No supervisor had authorized Caldwell to bypass that procedure.
No credible evidence suggested that Marcus had threatened anyone.
The gold badge was authentic.
The uniform was authentic.
The briefcase contained appointment documents, a copy of the city's official introduction letter, and administrative materials prepared for his first day.
The officers had not opened the briefcase before handcuffing him.
They had not contacted the command office.
They had not asked dispatch to verify his name until the arrival announcement came through.
They had not followed the procedure that existed precisely for situations in which an officer questioned someone's identity.
Dr. Price summarized the preliminary findings in a meeting with the oversight team.
"The detention was unsupported by the information available at the time. The force used against Chief Reed appears inconsistent with the circumstances documented in the video. The officers had a clear verification option and did not use it."
The city attorney asked whether the team had established racial discrimination.
"Not yet," Dr. Price replied. "We have evidence of unjustified assumptions and procedural failures. Determining whether race influenced the decisions requires additional evidence and broader analysis."
Marcus appreciated the distinction.
He had no interest in a conclusion that went beyond what could be demonstrated.
But he also knew that refusing to investigate a possibility because it was uncomfortable would be another kind of failure.
On the third day, the digital-forensics specialist discovered something unusual.
A group of complaint records had been exported from the department's database several months earlier.
The exports were not inherently suspicious. Administrative personnel routinely generated reports for legal counsel, budget reviews, and performance meetings.
What attracted the specialist's attention was the difference between the exported records and the current database.
Several cases had been closed after the exports were created.
That alone was not necessarily improper.
But the audit histories showed that some classifications had changed shortly before closure.
A complaint initially categorized as a potential use-of-force violation had become a "communication concern."
A complaint alleging retaliation had been reclassified as an "employee relationship issue."
Another, involving missing evidence, had been marked "administratively resolved" without a recorded explanation of what had been resolved.
The specialist compared the dates with the department's reporting calendar.
In each case, the reclassification occurred shortly before a monthly performance report was submitted.
Dr. Price read the preliminary analysis twice.
"Could these changes be legitimate?"
"Some could be," the specialist said. "The problem is that the audit history doesn't show consistent supporting documentation. We need to determine whether the changes reflected new evidence or simply changed the appearance of the numbers."
The team requested the original case files.
Several arrived within hours.
Others were delayed.
A few could not be located in the expected folders.
When investigators asked why, they received different explanations.
One employee said the files had been moved during a software migration.
Another said they were stored in a restricted archive.
A third insisted that the records had never existed in the format requested.
Dr. Price instructed her team not to treat missing documents as proof of deliberate destruction.
"Absence can have several explanations," she reminded them. "We document what we know and continue looking."
The search led to a shared drive used by Administrative Services.
Inside an old directory labeled QUALITY REVIEW—ARCHIVE, the digital specialist found a spreadsheet that had not been included in the department's official reports.
Its filename was Closure Alignment—Internal Use Only.
The spreadsheet contained hundreds of complaint numbers, dates, classification changes, and short notes.
Some entries were routine.
Others were troubling.
One column was labeled PREFERRED DISPOSITION.
Another was labeled REPORTING IMPACT.
A third contained abbreviated instructions such as:
Move to communication.
Close before monthly review.
Supervisor approval pending.
Avoid repeat escalation.
None of these phrases, by themselves, established criminal wrongdoing.
But the spreadsheet suggested that at least some complaint classifications might have been influenced by administrative objectives unrelated to the evidence.
The investigators traced the document's metadata.
It had been created by an account associated with Captain Paul Mercer.
The most recent modifications had occurred six weeks earlier.
Dr. Price requested an interview.
Mercer arrived wearing a navy suit and carrying a folder.
"I understand you've found a working spreadsheet," he said.
"We have," Dr. Price replied.
"That file was never an official policy document."
"Who created it?"
"Administrative Services personnel used several tools during the transition to the current reporting system."
"Did you create it?"
Mercer glanced at the printed copy.
"I may have contributed to it."
"What does 'preferred disposition' mean?"
"It was a planning field. Supervisors needed to identify the likely administrative outcome of a case."
"Why would the spreadsheet include 'reporting impact'?"
"To understand how open cases affected performance statistics."
"Did performance statistics influence the decision to close complaints?"
"No."
Dr. Price pointed to an entry.
"This complaint was classified as a potential use-of-force violation. Three days later, it was reclassified as a communication concern. The same entry says, 'Close before monthly review.' Can you explain that?"
Mercer studied the page.
"I would need to review the case file."
"The original file is missing from the expected location."
"Then I cannot give you an accurate answer."
"Did you instruct anyone to change the classification?"
"I don't recall doing so."
"Did you authorize the spreadsheet's use?"
"I authorized administrative staff to develop tracking tools."
"Did you review the results?"
"Of course."
"Did you question the number of complaints closed without sustained findings?"
Mercer shifted in his chair.
"The department receives many complaints that do not justify formal discipline."
"That was not my question."
"I reviewed the overall figures."
"Did you review the individual files?"
"Not every file."
"Then how did you know the classifications were accurate?"
Mercer looked toward the attorney seated beside him.
The attorney intervened.
"We would like to clarify the scope of this interview."
Dr. Price closed her notebook.
"That is your right. We will provide a written request for the relevant records and follow the established process."
Mercer rose.
Before leaving, he turned toward the investigators.
"I hope you understand that administrative decisions are not always as simple as they appear on paper."