Every investigator eventually gets the file that begins with a sentence like, “This happened a long time ago, but…” Sometimes it’s eight years. Sometimes it’s eighteen. A complaint surfaces about conduct that occurred far enough in that past that names may have changed, cities might have changed, and witnesses are no longer available. The events are old, but the obligation to investigate them properly is brand new.
I’ve come to think of these as the cold cases of our profession. And like many investigations, they carry a particular kind of difficulty, not because the truth is unknowable, but because time has been quietly working against you the whole while you weren’t looking.
In regulatory work especially, historical matters are becoming more common, not less. Mandatory reporting has expanded. People who once stayed silent are finding the confidence to come forward. Records that used to live in a filing cabinet now live forever on a server somewhere. The result is that a regulator, or an employer, can find itself examining conduct from a decade ago with the same expectation of rigour it would apply to something that happened last Tuesday. The challenge is real. It is also manageable, if you understand what the passage of time erodes
What time takes away
The first casualty is memory, and it’s worth being honest about how badly it degrades. Witnesses don’t just forget details over the years; they unconsciously revise them. They fill gaps with what they’ve since heard, read, or assumed. A complainant who has retold their account fifty times has, in a sense, rehearsed a version of events that may be sincerely believed and still drift from what happened. None of this means anyone is lying. It means the raw material you’re working with has aged, and you must handle it accordingly.
The second casualty is documentation. Email systems get migrated. Personnel files get purged under a retention policy. The supervisor who would have had the contemporaneous notes has retired and can’t be located. Physical evidence is gone. In a recent matter I worked on, the single most important document, the one that would have settled a factual dispute outright, had been destroyed entirely lawfully, three years before anyone knew there was a dispute to settle.
The third casualty is the respondent’s ability to answer. This is the one investigators most often overlook, and it’s the one that fairness hinges on. A person asked to account for a specific interaction from years ago is at a genuine disadvantage. They may have no records, no memory of the day in question, and no way to call the alibi witness who has since passed away. The longer the delay, the harder it becomes for them to mount a meaningful response, and a process that ignores that is building a finding on sand.
Delay is a fairness issue, not just an inconvenience
We tend to treat the age of a matter as a practical nuisance, involving more digging and more dead ends. But delay is fundamentally a procedural fairness issue, and it deserves to be named as one early in the file.
The question isn’t simply can we still investigate this? It’s can we investigate this in a way that’s fair to everyone, given what’s been lost? Those are different questions. Sometimes the honest answer is that you can reach a sound conclusion despite the years. Sometimes it’s that the passage of time has compromised the respondent’s ability to answer so badly that any adverse finding would be unsafe. A good investigator is willing to say so. Naming the limitation isn’t weakness; it’s the thing that protects the integrity of whatever conclusion you do reach.
How to overcome the challenges
So, what do you do with the cold case on your desk? A few things have served me well.
Reconstruct the timeline before you interview anyone. With historical matters, your most reliable witness is often the documentary record, what little of it survives, precisely because it doesn’t drift the way memory does. Build the chronology from whatever is fixed and dated: surviving emails, system logs, calendars, HR records, regulatory filings. Anchor the human accounts to those fixed points rather than the other way around. When a witness’s memory conflicts with a contemporaneous document, the document usually wins, and you want to know that before you’re sitting across from them.
Chase the records early and aggressively. Retention policies are running against you in real time. The day you open a historical file is the day you issue litigation holds, request preservation, and track down every system that might still hold something. What’s recoverable today may be gone next quarter. I’ve learned to treat the first week of an aged matter as a race.
Interview for the gaps, gently. When you sit down with witnesses, distinguish carefully between what they actually remember and what they’ve reconstructed. Ask how they know what they’re telling you. Probe the source of each recollection. And resist the temptation to read confidence as accuracy. In historical matters the two are often inversely related. The witness who says “I can’t be certain after all this time, but…” is frequently more reliable than the one with a cinematic memory of a single afternoon a decade ago.
Assess credibility without leaning on demeanour. You can’t fairly judge whether someone is truthful about a distant event by how nervous they look recounting it. Anyone would be hazy. Look instead for internal consistency, consistency with the surviving record, and inherent plausibility. That’s always good practice; with aged matters it’s essential.
Account for the delay in your findings (explicitly). When you write your report, don’t pretend the years didn’t happen. State what was lost, what couldn’t be tested, and how that affected your analysis. If the evidence is strong enough to support a conclusion anyway, your transparency makes that conclusion more defensible, not less. If it isn’t, you’ve documented exactly why, and you’ve done your job honestly.
The real test
Historical matters test something deeper than our research skills. They test our willingness to hold two ideas at once: that age is never a reason to dismiss a serious allegation, and that age can genuinely compromise our ability to prove one. Both are true. Leaning entirely on the first turns us into rubber stamps for the passage of time; leaning entirely on the second allows misconduct hide behind the calendar.
The investigator’s task is to sit in that tension and work it honestly, to dig hard, account for what’s been lost, and reach the soundest conclusion the surviving evidence will bear. Not the conclusion the complainant wants, or the one the respondent hopes for, but the one the record can actually support.
The case is cold. The standard is not.
