By Cortni Lawson, Founder & CEO, InfraNet HR
Cross-domain patterns that signal legal risk rarely announce themselves in a single complaint. They show up as separate events — a comp claim, a leave request, a safety report, a performance write-up — that only look dangerous when you see them together.
You probably know that retaliation is illegal. You probably have a policy about it. You probably think your company wouldn't do it.
But here's what keeps compliance teams up at night: retaliation doesn't always look like what you think it looks like.
It doesn't usually look like, "We fired you because you filed a workers' comp claim."
It looks like a series of separate decisions that, when you see them together, add up to retaliation.
What Patterns Actually Look Like
Let me describe some real patterns. These are from actual cases, simplified but accurate.
Pattern 1: The Sudden Discipline
Employee files a workers' comp claim for a back injury. Claim is approved. Employee takes medical leave as instructed. Six weeks later, employee returns to modified duty.
Two weeks after return, employee gets written up for a mistake that happened three months ago. Nobody mentioned it at the time. But now, suddenly, it's documented. Now it's formal. Now it's a problem.
One write-up? Could be coincidental.
But if three employees who filed comp claims all got written up for old issues shortly after returning, that's not coincidence. That's a pattern.
And that pattern is actionable liability.
Pattern 2: The Accommodation Breakdown
Employee requests an ADA accommodation for a disability disclosed during onboarding. HR approves it. Accommodation is documented. Manager is supposed to implement it.
Three weeks later, the employee is struggling. The accommodation either wasn't implemented or was implemented half-heartedly. Employee complains to HR. HR talks to the manager, who says the accommodation is "too disruptive" or "slowing down the team."
HR tries to problem-solve, but the accommodation never gets properly implemented. Employee, frustrated, files a complaint with HR or goes to the EEOC.
One accommodation failure? Mistakes happen. But if five employees with documented accommodations all experienced implementation failures, and three of them eventually filed complaints or quit, that's a pattern.
And that pattern suggests a compliance failure across the board.
Pattern 3: The Leave-to-Termination Pipeline
Employee requests FMLA leave. It's approved. Employee takes the leave as certified.
When the employee returns, things are different. Their schedule has changed. Their manager seems cold. A project they were working on got reassigned. It feels like they're being pushed out.
Within 90 days of return, they're terminated. The documented reason is unrelated to the leave.
One leave-to-termination case? Could be legitimate. Employees leave after time off all the time.
But if you notice that 20% of employees who take FMLA leave are terminated within a year, and the general termination rate is 10%, you have a pattern. And that pattern is FMLA interference.
Pattern 4: The Safety Complaint Silence
Employee reports a safety hazard. It's documented in the safety system. The hazard is serious.
But then nothing happens. The hazard isn't fixed. Corrective action isn't taken. Follow-up doesn't occur.
Meanwhile, that employee is reassigned. Their hours are cut. They get written up for unrelated issues.
One safety complaint with no follow-up? Overworked. Multiple safety complaints, none of which get addressed, all followed by adverse action against the reporter? That's a pattern of retaliation against safety reporters.
That's OSHA liability.
Pattern 5: The Concentrated Manager
You look at one manager's history. Over 18 months, this manager has:
- Had five employees file safety complaints
- Had four workers' comp claims
- Had three ADA accommodations requested
- Terminated six employees
That's not unusual in itself. Managers manage.
But you notice something: every employee this manager supervised who filed any kind of complaint or protected activity got terminated within 12 months. The general termination rate in the company is 8%. For this manager, it's 40%.
That's not coincidence. That's a manager who retaliates against complaints.
And that's a company liability, because now you know about it.
Why These Patterns Are Invisible in Fragmented Systems
Each of these patterns requires looking across multiple systems.
The sudden discipline pattern requires correlating workers' comp claims with performance documentation. Two different systems. Two different departments sometimes.
The accommodation breakdown requires seeing the accommodation request, the implementation plan, the employee complaint, and the employment status change. Four different systems.
The leave-to-termination pattern requires tracking FMLA approvals, leave taken, return dates, and subsequent employment actions. Again, multiple systems.
The safety complaint silence pattern requires seeing safety complaints, follow-up actions, and employment actions. Three systems.
The concentrated manager pattern requires pulling termination data, complaint data, accommodation requests, and safety reports, and then analyzing them by manager.
Your safety system isn't asking, "Did anyone who reported a safety hazard get terminated?"
Your leave system isn't asking, "Do employees who take FMLA leave have higher termination rates?"
Your workers' comp system isn't asking, "Are employees who file claims getting written up shortly after returning?"
These questions require a single view of employee data across domains. They require correlation. They require analysis.
And they're invisible when data is scattered.
The Legal Standard for "Should Have Known"
Here's where it gets legally interesting: if you have access to data, and the pattern is there, but you're not looking for it, that's negligence.
It's worse than negligence. It's deliberate indifference.
The argument goes: "You had all the information. You were just not analyzing it. That demonstrates a compliance program that doesn't work."
And the opposing counsel is right.
The DOJ Compliance Program Guidance specifically asks: do you monitor for compliance risks? Do you analyze data to identify patterns? Do you have controls in place to detect violations?
If you're not correlating data across systems, you're not monitoring. You're not analyzing. You don't have controls.
You have documentation. You have isolation. You don't have a compliance program.
What Correlation Actually Means
This is the part that sounds complicated but isn't.
Correlation doesn't mean you've determined someone violated the law. It means you've noticed something worth investigating.
Example: You notice that employees who file safety complaints have a 30% termination rate, while employees who don't file complaints have a 5% termination rate.
That correlation doesn't prove retaliation. It proves you should investigate. Maybe the employees who file complaints are problem employees. Maybe there's a legitimate reason for the disparity.
But if you're not asking the question, you'll never know.
And if the EEOC asks you, "Did you analyze termination rates by protected activity?" and the answer is "no, we don't correlate that data," you're admitting your compliance program doesn't work.
What You Should Be Asking
If you want to know whether you have retaliation risk, you should be asking questions like:
On workers' comp:
- "How many employees who file comp claims get adverse action within 90 days of return?"
- "Is that rate different from employees who don't file claims?"
- "Which managers have the highest rate of adverse action against returning employees?"
On FMLA:
- "How many employees who take FMLA leave get terminated within 12 months?"
- "Is that rate different from employees who don't take FMLA leave?"
- "Are there any managers or departments with unusual rates?"
On safety:
- "How many safety complaints result in corrective action?"
- "How many employees who file safety complaints get adverse action within 90 days?"
- "Which departments have patterns of unreported hazards?"
On accommodations:
- "How many accommodation requests get implemented as approved?"
- "How many employees with accommodations get terminated?"
- "Are there patterns by manager or department?"
Across domains:
- "Are there employees with multiple types of protected activity (comp claim + FMLA + safety complaint)?"
- "Did those employees get adverse action?"
- "What's the pattern?"
These questions are not theoretical. They're practical. And they're the questions the DOJ, the EEOC, and plaintiff attorneys will ask if something goes wrong.
If you can answer them because you analyze your data, great. You have a compliance program.
If you can't answer them because you're not correlating data across systems, you have a documentation problem. You have silos. You don't have a program.
The Conversation with Your Attorney
Your attorney asks: "Do you monitor for retaliation risk?"
You say: "Yes, we have a strong investigation process and we document everything."
Your attorney follows up: "Do you analyze data to identify patterns of retaliation?"
You pause. "Well, we… we have systems for investigations, and systems for workers' comp, and systems for leave…"
Your attorney knows what that means. It means no. You don't analyze across systems. You document within systems.
Now, if you get sued, your attorney has to explain why you didn't have a program that could identify a pattern that's obvious in hindsight.
That's a bad conversation to have.
A better conversation:
Your attorney asks: "Do you monitor for retaliation risk?"
You say: "Yes. We track protected activities — workers' comp filings, FMLA requests, safety complaints, accommodations, discrimination complaints. We correlate them with employment actions. We analyze for patterns. We escalate when we see concerning correlations."
Now you have a program.
That's the difference correlation makes.