She went back to her desk that night and opened her timeline again. She added a new column, one she had not needed before. It was headed "Response." In it, she began to log each delay, each canceled meeting, each vague warning. She dated every entry. She kept the language neutral. She did not speculate about motive. She simply recorded what had happened, when it had happened, and who had been present.
If anyone ever asked her later why she had continued, she wanted the answer to be in the record, clear and unarguable: because the questions were legitimate, the documents were real, and the people who had lost their money deserved someone willing to ask.
Chapter 8: The Informants
The names of confidential informants were not supposed to appear in ordinary incident reports. They were usually reduced to a number, a code, or a vague phrase: a reliable source, a cooperating individual, a person familiar with the area. Maya had seen those phrases hundreds of times. They were designed to protect people who provided information at real risk to themselves, and in legitimate investigations, that protection mattered.
In the forty-one reports, however, the protection had failed in a strange and revealing way.
Maya noticed it first in the details. A report from March described a stop on a road near the north side of Riverton. The officer wrote that he had received information that the driver might be carrying a large amount of cash. The source was listed only as a confidential informant. A report from May described a stop on a different road, in a different neighborhood, with a different driver. The wording was almost identical. A report from June used the same phrasing again. The same sentence structure. The same justification. The same conclusion.
Maya placed the three reports side by side on her desk. She read them slowly, then read them again. The language was too similar to be coincidence. It was the language of a template, copied and pasted, with only the dates, names, and amounts changed.
She began to track the informants as if they were characters in a case, not shadows. She gave each recurring pattern a label. Informant A appeared in reports where the stop was justified by a tip about cash in the vehicle. Informant B appeared in reports where the stop was justified by a tip about a planned exchange. Informant C appeared in reports where the officer wrote that he had observed the driver meeting briefly with another person in a parking lot.
She did not know their real names. She did not try to learn them. That was not her job, and she understood the danger of exposing someone who might genuinely be at risk. What interested her was not who they were, but how often they appeared, and how perfectly their tips seemed to predict a seizure.
Over the course of the forty-one reports, Informant A was credited in fourteen stops. Informant B was credited in eleven. Informant C was credited in nine. The remaining seven reports cited no informant at all, or cited a source so vaguely described that Maya could not categorize it. The numbers troubled her. In a city the size of Riverton, it was possible for a small number of informants to be productive. It was also possible for a small number of informants to be invented, reused, or pressured into providing the same story whenever an officer needed a reason for a stop.
She checked the dates. Informant A was credited with tips on days when the same officer was listed as working in two different parts of the city. Informant B was credited with information about drivers who had no prior contact with the police, no record, no connection to any open investigation that Maya could find. Informant C was credited with tips that led to seizures, but never to arrests, never to charges that would have required the informant to appear in court or to be tested under cross-examination.
Maya wrote those observations down, carefully, without drawing a conclusion she could not support. She was building a circumstantial case, and she knew that circumstantial cases required discipline. Every inference had to be tied to a document. Every document had to be preserved.
She took her findings to Samuel Osei, her defense attorney and former law-school classmate. They met in his office after hours, when the hallway was empty and the phones had stopped ringing. Samuel was a careful listener, the kind of lawyer who asked a question only after he had heard the full answer to the previous one. Maya laid out the reports in groups, explained her labeling system, and showed him the overlaps.
Samuel studied the pages for a long time. "If these informants are real," he said at last, "then the department should be able to produce records of how they were recruited, how they were paid, and how their information was verified. If they can't produce those records, or if the records don't match the reports, then you have a problem for the department, not just a question."
"And if they are not real?" Maya asked.
"Then you have a bigger problem," Samuel said. "Because then the stops were not based on tips at all. They were based on something else. And the reports were written to hide that."
Maya nodded. She had reached the same conclusion, but she wanted to hear it from someone else, someone who would challenge her if she was reaching too far.
She also consulted a former prosecutor she trusted, a woman who had left the office years earlier and now taught criminal procedure. They spoke by phone, without names, without case numbers. Maya described the pattern in general terms. The former prosecutor was quiet for a moment, then said, "In my experience, when the same informant shows up in that many cases without ever being disclosed, one of three things is happening. The informant is being overused. The informant is being protected because the cases are still active. Or the informant is a convenient explanation. You need to figure out which one it is, and you need independent evidence, not just the reports themselves."
Maya understood the warning. The reports alone could show repetition. They could not prove fabrication. For that, she would need something outside the reports: payment records, handler logs, dispatch audio, or testimony from someone inside the system who had seen how the tips were generated.
She returned to the ledger. She checked whether any payments to informants were recorded. They were not. The ledger tracked seizures, not informant compensation. She checked the photographs. None of them showed an informant. She checked the reports again for any mention of a meeting, a phone call, a payment. There was nothing. The informants existed only as phrases on paper, appearing when needed, disappearing when the stop was complete.
That absence began to feel as significant as the repetition.
Maya added a new section to her timeline. She called it "Informant Activity," and she plotted each credited tip by date and officer. The chart showed clusters. On certain weeks, Informant A appeared three times in four days. On other weeks, no informant was mentioned at all. The clusters lined up with periods when seizures were highest. When seizures dropped, the informants went quiet. When seizures rose, the informants became active again.
She stared at the chart for a long time. Correlation was not causation. She reminded herself of that, the way she had reminded junior associates a dozen times. But correlation, combined with identical language, combined with a lack of corroborating records, combined with the concentration of stops in the same neighborhoods, was beginning to form a shape she recognized from her civil-rights work. It was the shape of a practice that had become routine, a routine that had become policy, and a policy that no one had bothered to write down because writing it down would have made it visible.