Collins stopped.
She was right.
He had spent years believing that a clear chain of command made the department efficient. Now he was discovering that the same chain could create the appearance—or the reality—of pressure when the people at the top wanted answers.
"Fine," he said. "Put the request through the review team. I won't contact them directly."
After ending the call, he opened the department's archive and searched for the memorandum.
The search returned a list of related documents, including meeting notes from a policy committee held three years earlier.
One entry caught his attention.
Complaint Resolution Efficiency Initiative — Final Review.
He opened the file.
The initiative had been created after the department experienced a backlog of complaints. The stated goal was to reduce delays, standardize reviews, and ensure that officers were not left under investigation indefinitely.
The goals sounded reasonable.
The performance targets were more troubling.
Reviewers had been encouraged to close cases within a fixed number of days. Supervisors were evaluated partly on how quickly their units resolved complaints. Cases that required additional interviews, external records, or technical analysis were described as exceptions that should be avoided when possible.
Collins read the final page.
A table showed that the department had significantly reduced the average time required to close complaints after the initiative began.
The table did not show how many complaints had been reopened.
It did not show how often complainants had been interviewed.
It did not show how many cases had been closed before all available evidence was collected.
The department had measured speed.
It had not adequately measured accuracy.
Collins closed the document and sat motionless.
He could already imagine the explanation that would be offered. The backlog had been real. The pressure had been real. The intention had been to improve the process.
But a system could be designed to achieve a target while failing the purpose for which the target existed.
If reviewers were rewarded for closing complaints quickly, and if officers' reports were treated as the default account, the result might be predictable even without anyone explicitly instructing investigators to ignore the truth.
The question was no longer merely whether a particular officer had acted improperly.
It was whether the department had built a process that made improper conduct easier to overlook.
And if it had, who would have the courage to admit it?
That afternoon, the independent review team requested access to the original policy files, the complaint database, and the records of cases closed under the initiative.
The request went through the city attorney's office.
Within an hour, Commissioner Collins approved it.
He also issued an order prohibiting the destruction or alteration of any relevant records and directing supervisors to preserve emails, internal messages, and archived documents connected to complaint reviews.
The order did not announce wrongdoing.
It did not promise a sweeping reform.
It simply required the department to stop treating the past as something that could be left unexamined.
For the people whose complaints had been dismissed, that small procedural step could mean the difference between a closed file and a second look.
For the officers whose reports were under scrutiny, it meant that the records would be examined in their original form.
And for the department itself, it meant that the questions raised in Marcus Reed's case would no longer remain confined to one courtroom.
Two days later, Angela received another message from the city attorney.
The original memorandum had been located.
It included handwritten notes from a policy committee meeting, along with an email chain discussing the proposed language.
Angela read the email chain carefully.
One message came from a senior administrator who had warned that the proposed presumption could discourage reviewers from pursuing complaints supported by recordings that were incomplete but still relevant.
Another message dismissed the concern.
The response argued that a more demanding review process would increase costs, prolong investigations, and expose the department to unnecessary litigation.
The final version of the memorandum had retained the presumption in favor of the officer's written account.
Angela printed the documents and carried them to Rachel's office.
"Look at this," she said.
Rachel read the exchange.
"Who is the administrator who raised the objection?"
"Deputy Commissioner Elaine Foster."
"Is she still with the department?"
"She retired last year."
Rachel looked at the date.
"Then we need to speak with her."
Angela nodded.
"And we need to find out whether she was overruled, whether her concerns were documented elsewhere, and whether the final policy was implemented as written."
Rachel closed the folder.
"This could change the direction of the case."
"It could also create a problem if we overstate what it proves."
"Agreed. The emails show that someone raised concerns. They don't establish why every later complaint was closed."
"Exactly. We have to follow the evidence, not jump to the conclusion we expect."
Rachel smiled faintly.
"You're reminding me to do the job properly."
"I'm reminding both of us."
They contacted Elaine Foster through the appropriate legal channels.
She agreed to meet with them the following week.
When the meeting began, Foster sat across from Angela and Rachel in a small conference room. She was in her early sixties, with silver hair and a careful, measured manner.
"I wondered when someone would ask about that memorandum," she said.
Rachel opened her notebook.
"Did you write the email raising concerns about the proposed presumption?"
"Yes."
"What concerned you?"
Foster folded her hands.
"The original draft said that reviewers should examine the totality of the evidence before reaching a conclusion. The later draft gave the officer's report a preferred status. I believed that was a mistake."
"Why?"
"Because an officer's report is evidence. It is not the investigation itself."
Angela nodded.
"Did you express that concern at the committee meeting?"
"Yes."
"What happened?"
"Some members agreed with me. Others argued that the department needed a consistent standard to protect officers from complaints that could not be substantiated."
"Was that a legitimate concern?"
"Of course. Officers deserve fair investigations too. A false complaint can damage a person's reputation and career. But protecting officers from unfair allegations is not the same as treating their accounts as more credible by default."
Rachel wrote down the statement.