The following morning, Morales received the results of a preliminary search of electronic communications related to the old complaints.
The search had been ordered after the audit identified unexplained closures in several cases. Investigators were looking for messages that might clarify why certain files had been marked incomplete or removed from the active queue.
Most of the results were ordinary administrative exchanges. Supervisors asked for missing forms, investigators requested extensions, and clerks reminded officers about overdue reports.
One exchange, however, stood out.
It involved a group of supervisors discussing the backlog in a district office. The messages were several years old and did not refer to a particular resident. They focused on the pressure to reduce the number of unresolved complaints before an internal performance review.
One supervisor wrote that the district needed to “get the numbers under control.”
Another responded that files could be moved out of the active queue when the complainant had not provided enough information.
A third message warned that this approach could create problems if the complainant had never been properly contacted.
The discussion ended without a clear resolution.
Morales read the exchange twice.
It did not establish that anyone had deliberately concealed misconduct. The messages could reflect a genuine disagreement about how to manage incomplete files. But they raised a new question: had the pressure to improve performance figures influenced the way complaints were classified?
She brought the messages to the independent auditors.
“Can we identify who had authority over the process at that time?” she asked.
Whitfield examined the names and dates.
“We can establish the reporting structure. We will need to determine whether the messages reflect an actual practice or only a discussion about a possible approach.”
“Can we compare the dates with the complaint closures?”
“Yes. If the same period shows an unusual increase in files marked incomplete, that would be relevant. It would not prove intent by itself, but it could guide the next stage of review.”
Morales approved the comparison.
Within days, the analysts identified a sharp increase in incomplete classifications during the month before the performance review. The increase did not appear in every district, and there were plausible explanations involving staffing changes and system migration. But the pattern was significant enough to require further examination.
Commissioner Shaw was informed immediately.
She asked whether the evidence supported a conclusion that supervisors had manipulated complaint statistics.
“Not yet,” Morales replied. “It supports further investigation into whether performance pressure affected classification decisions. We need to establish who made those decisions, what guidance they received, and whether the same rules were applied consistently.”
Shaw looked at the printed messages.
“Could the problem have been caused by a policy that rewarded closure rates without measuring the quality of the investigation?”
“Yes. That is one possibility.”
“Then we need to examine the incentives, not just the individuals.”
Morales agreed.
The department's performance system had rewarded units for clearing backlogs and meeting processing targets. Those goals were not inherently improper. Delayed complaints could leave residents waiting for answers and officers uncertain about allegations against them. But if the system rewarded closure without equally measuring investigative quality, it could encourage decisions that made the numbers look better without resolving the underlying questions.
The department had not necessarily instructed supervisors to close complaints improperly. Yet it might have created conditions in which a quick administrative closure appeared more valuable than a thorough investigation.
That distinction became central to the next phase of the audit.
Marcus reviewed the findings with Shaw late one evening.
“If we only discipline individuals, we may miss the reason the practice developed,” he said.
“And if we focus entirely on the system, we risk allowing individuals to avoid responsibility for their decisions,” Shaw replied.
“Exactly.”
“So how do we balance it?”
“We investigate both. We determine whether specific people violated their obligations, and we determine whether the department created incentives or procedures that made those failures more likely.”
Shaw looked tired.
“Some will say that means nobody is responsible because everyone contributed.”
“Then we must be precise. A system can influence behavior without eliminating personal responsibility. And personal responsibility doesn't mean the system should remain unchanged.”
Shaw closed the report.
“I want the final recommendations to include measurable standards. Not just new language in a policy manual.”
Marcus nodded.
“We can measure whether complainants receive confirmation, whether witnesses are contacted when identified, whether closures include evidence-based explanations, and whether supervisors complete the required reviews.”
“And who checks the measurements?”
“An independent oversight body, at least during the first phase.”
“Will the department agree to that?”