The man eventually produced a building access card. Mitchell examined it, then asked for a driver's license. The man asked why the access card was insufficient to establish that he lived there.
The conversation continued for several minutes. No arrest occurred. No weapon was visible. The man did not threaten Mitchell or attempt to enter a restricted area. At the end, Mitchell returned the card and told him to "make better choices" about how he spent his time.
The man looked confused.
"I was buying groceries," he said.
Mitchell turned away without answering.
Harris watched the final seconds in silence.
The recording did not establish every detail of the earlier complaint, but it confirmed that the encounter contained behavior similar to what Danielle had described. Mitchell had demanded identification, failed to explain a clear legal basis, and treated ordinary questions as signs of noncooperation.
The question now was whether the earlier complaint had been reviewed in light of the recording.
Harris opened the electronic case file and read the closing summary again.
The summary stated that the officer had conducted a brief consensual inquiry and that the individual had been released without further action. It concluded that no policy violation was identified.
But the video raised an obvious question: had the reviewer considered whether the encounter had remained consensual after Mitchell began demanding identification and framing questions as refusal to cooperate?
The summary did not say.
Harris called the records administrator.
"Can we establish when this recording was first accessed?"
"We have a sequence of system events, but not a reliable record identifying the person who viewed the video. The old platform recorded some actions differently from the current system."
"Could we recover the original audit database?"
"The system administrator believes a backup may exist. It will take additional work to locate and verify it."
"Please proceed. Don't assume that the absence of a user name means no one viewed it. We need to know what the records can and cannot establish."
"Understood."
Harris then requested the full dispatch audio and the original report from the earlier encounter.
When he finished, he sat back and looked at the screen.
One recording did not establish that Mitchell had violated policy in every similar situation. It did not prove that the earlier complaint had been deliberately mishandled. But it demonstrated that the complaint summary had not captured the full character of the encounter, and it raised legitimate questions about the basis for the conclusion that no violation had occurred.
Harris saved his notes and marked the video for review by the department's legal adviser.
He was no longer investigating only what had happened on Danielle's porch.
He was examining whether the department had repeatedly encountered warning signs and failed to respond to them with sufficient care.
Later that afternoon, the legal adviser sent Harris a preliminary assessment of the retention issue.
The assessment explained that recordings associated with formal complaints could be subject to a longer preservation period than recordings from ordinary calls, depending on the classification of the complaint and the policy in force at the time. The documents available so far did not establish whether the expired recording had been classified correctly.
Harris added the issue to the investigation.
He also asked for a list of all complaints involving Mitchell during the previous five years, including complaints that had been withdrawn, categorized as informal inquiries, or closed without a finding. He wanted to compare the initial allegations with the final dispositions and determine whether similar patterns had been recorded under different categories.
That request generated an immediate response from the administrative unit.
There were more records than the team had initially identified.
Some were formal civilian complaints. Others were supervisor notes documenting concerns raised during routine interactions. A few were entries in which a person had called to ask whether an officer's behavior was permitted but had not completed a formal complaint form.
The entries did not all concern the same type of conduct. Some involved tone of voice, some involved prolonged questioning, and others concerned disputes over identification. Several had been resolved without any formal finding.
Harris knew that the number alone could be misleading. An officer who worked many public-facing shifts might receive more complaints simply because of the volume of encounters. Each record needed to be assessed on its own facts.
But he also knew that a department could miss a meaningful pattern if it treated every concern as an unrelated event.
He asked the administrative unit to create a chronology and include the basis for each disposition.
The first draft arrived the following morning.
Harris read it slowly.
A complaint from three years earlier described Mitchell questioning a woman outside a community center after she told him she was waiting for her son. The report said the woman had been "uncooperative." A supervisor's note stated that the matter had been resolved after the officer explained his concerns, but there was no detailed record of what those concerns had been.