“Refusal to Work” Means Non‑Performance, Not Safer Performance:
Commentary on Berberich v. Kansas City Southern Railway Company and FRSA § 20109(b)(1)(B)

I. Introduction

In Berberich v. Kansas City Southern Railway Company, No. 24‑3154 (10th Cir. Dec. 22, 2025) (published), the Tenth Circuit issued a significant interpretation of the “refusal to work” retaliation protection contained in the Federal Railroad Safety Act (FRSA), 49 U.S.C. § 20109(b)(1)(B).

The plaintiff, Justin Berberich, a conductor at Kansas City Southern Railway Company’s (KCSR) Knoche Yard in Kansas City, Missouri, claimed he was terminated in retaliation for engaging in protected activity under the FRSA. The only alleged protected activity remaining on appeal was his decision in January 2019 to walk a substantial distance and personally throw a rail switch—rather than allow the locomotive engineer to do so—because he believed having the engineer leave the controls created a serious safety risk.

The Tenth Circuit affirmed summary judgment for KCSR, but on a different ground than the district court. Without reaching whether the decisionmaker knew of the alleged protected activity, the court held that Berberich had not engaged in any protected activity under § 20109(b)(1)(B) at all. The decision turns on two core statutory interpretations:

• What it means to “refus[e] to work” under § 20109(b)(1)(B); and
• What qualifies as being “confronted by a hazardous safety or security condition” involving an “imminent danger of death or serious injury” under § 20109(b)(2).

The court’s ruling narrows the circumstances in which railroad employees can claim FRSA retaliation based on alleged refusals to work in dangerous conditions. It also illustrates important evidentiary and appellate‑practice principles that will shape future FRSA litigation in the Tenth Circuit and beyond.

II. Summary of the Opinion

A. Procedural Posture

After his termination, Berberich:

1. Filed a complaint with OSHA alleging retaliation for reporting safety concerns, reporting injuries, and insisting on following safety rules.
2. Lost before an administrative law judge (ALJ), who found no protected activity and no retaliatory motive.
3. Filed a de novo FRSA retaliation action in the District of Kansas after the Secretary of Labor failed to issue a final decision within 210 days, as permitted by § 20109(d)(3).

In district court, he asserted three separate protected activities; only one survived to the appeal: his alleged refusal in January 2019 to follow a supposed “standing order” that engineers, not conductors, throw the switches when that would expedite train movements.

The district court granted summary judgment for KCSR on the ground that there was insufficient evidence that the ultimate decisionmaker, General Manager Chad Devenney, knew of this alleged protected activity—thus knocking out the “employer knowledge” element of a prima facie FRSA claim.

B. Tenth Circuit’s Disposition

The Tenth Circuit affirmed, but on an alternative ground:

• It did not decide whether Devenney knew about the January 2019 incident (and thus did not decide the “knowledge” or “cat’s paw” issues).
• Instead, it held there was a “complete failure of proof” on the essential threshold element: Berberich had not engaged in protected activity under § 20109(b)(1)(B) because:
  • He did not “refus[e] to work” within the plain meaning of the statute; and
  • He was not confronted with a qualifying “hazardous safety condition” presenting an “imminent danger” when he lined the switch.

The panel further noted a third problem: there was no evidence of any true “standing order” compelling engineers to line switches in all circumstances. At most, the evidence showed occasional, ad hoc instructions or preferences from a supervisor (Pollard) to expedite moves.

Because this failure went to “protected activity,” an essential element of a FRSA retaliation claim, summary judgment was appropriate, and the court did not reach the remaining elements (knowledge, adverse action, causation) or KCSR’s possible affirmative defenses.

III. Factual and Legal Background

A. The Role of Conductors and Switches in Knoche Yard

As a conductor at KCSR’s Knoche Yard, Berberich’s duties included:

• Working primarily at the rear of trains to assemble (“make up”) and disassemble trains;
• Connecting railcars; and
• “Throwing” or “lining” mechanical switches to guide trains from one track to another.

The court cited a U.S. Department of Transportation report to describe how switches work and their role in moving and classifying rail cars in a yard. Some switches are manual (“hand‑thrown”), and they must be “lined” to align with the appropriate track.

A practical tension underlay the dispute: when the conductor is far back on a long train and the switch needing to be lined lies near the head end, having the conductor walk up and back can significantly delay movements. According to Berberich, yard supervisors sometimes tried to expedite operations by directing engineers (who are in the front cab) to get down from the locomotive and line the switch themselves.

B. Alleged “Standing Order” and Safety Concern

Berberich’s central narrative was:

• There was a “standing order” at Knoche Yard requiring engineers to throw switches in some circumstances, to expedite moves.
• He believed this was unsafe because engineers would have to leave the controls unattended while the train remained on air brakes only, contrary to KCSR’s own rule not to depend solely on air brakes when equipment is left unattended.
• On a January 2019 day, he “refused” to let the engineer throw a switch and instead walked the length of the train and threw it himself. He characterized this as a “refusal to work” in the face of a hazardous condition.

The record, however, undercut crucial elements of this story:

• The term “standing order” was introduced by counsel; Berberich himself did not originate the phrase.
• On cross‑examination, he clarified that there was no written directive or fixed rule; instead, supervisor Mike Pollard would sometimes, in particular situations, radio engineers to line switches when conductors were far back, in order to “expedite movements.”
• Another witness (engineer Jeffrey Spigarelli) corroborated only that Pollard occasionally gave such ad hoc instructions over the radio.

Thus, the court (and previously the ALJ) found no evidence of a true “standing order” always in effect, as opposed to sporadic managerial direction in specific operational scenarios.

C. The January 2019 Switch‑Lining Incident

On the critical day, Berberich was assembling a train and needed a switch ahead of the locomotive to be lined:

• The engineer radioed that a switch ahead of the train was lined against him.
• Berberich testified that, as conductor, it was his job to line the switch. He:
  • Called for a yard cab to transport him to the head end (none was available);
  • Then walked to the head end, lined the switch himself, and walked back to continue his work.
• He did not testify that the engineer was about to line the switch or had been instructed to do so on that particular occasion.
• He also testified he did not tell the engineer not to line the switch; rather, he relied on the “common practice” that the conductor lines switches.

Shortly after he returned from lining the switch, a manager confronted him and said he was “going to get [him] for delaying the trains,” allegedly at Pollard’s behest. But there was no testimony specifying that the delay was caused by his decision to line the switch rather than having the engineer do it.

D. The Termination and Alleged Retaliation

Approximately six weeks later, in February 2019:

• Pollard observed Berberich leaning on a rail car for about 40 seconds, a violation of KCSR’s safety rules.
• Following a disciplinary hearing, General Manager Devenney terminated him, allegedly pursuant to KCSR’s disciplinary matrix and his prior infractions.

Berberich alleged:

• He was being “bird dogged”—closely watched—for purposes of finding grounds to fire him after the January 2019 incident.
• Devenney knew about his January 2019 safety‑related conduct, as evidenced by:
  • Spigarelli’s testimony that he overheard Devenney on the phone with Pollard saying “We need to do something about [Berberich]” and that he was tired of hearing Berberich’s name associated with trains “on the law” (i.e., trains reaching the 12‑hour cap under the Hours of Service Act);
  • The temporal proximity (about six weeks) between the incident and termination; and
  • Alleged disproportionality in the punishment for leaning on the car.

A Public Law Board later found that while he had violated safety rules, dismissal was too severe and ordered his reinstatement without backpay. Separately, OSHA dismissed his FRSA complaint, and the ALJ rejected his claims on multiple grounds, including that:

• There was no credible evidence of the January 2019 incident as characterized by Berberich;
• No proven “standing order” existed requiring engineers to always throw switches; and
• Even under his version of events, he did not “refuse to work.”

IV. Legal Framework and the Court’s Core Holdings

A. FRSA Retaliation Under § 20109(b)(1)(B)

FRSA § 20109 is a whistleblower protection statute for railroad employees. It prohibits rail carriers from discriminating against employees for engaging in various forms of protected activity. Two subsections are especially relevant:

• § 20109(a)(2) – protects employees for “refus[ing] to violate or assist in the violation of any Federal law, rule, or regulation relating to railroad safety or security.”
• § 20109(b)(1)(B) – at issue here, protecting an employee for “refusing to work when confronted by a hazardous safety or security condition related to the performance of the employee’s duties,” provided that the stricter requirements of § 20109(b)(2) are met.

Under Tenth Circuit precedent (Lincoln v. BNSF Ry. Co., 900 F.3d 1166 (10th Cir. 2018)), a FRSA retaliation plaintiff must make a prima facie showing that:

1. The employee engaged in protected activity;
2. The employer knew of the protected activity;
3. The employee suffered an unfavorable personnel action; and
4. The protected activity was a contributing factor in the unfavorable personnel action.

The Berberich court focused exclusively on the first element—protected activity—because a failure on any essential element is fatal under Celotex Corp. v. Catrett, 477 U.S. 317 (1986).

B. Statutory Preconditions for a Protected “Refusal to Work”

Section 20109(b)(2) imposes three detailed conditions for a refusal to qualify as protected activity:

1. Good faith and no reasonable alternative:
   “The refusal is made in good faith and no reasonable alternative to the refusal is available to the employee.”
2. Objective reasonableness and imminence:
   A reasonable person in the employee’s circumstances would conclude that:
   • The hazardous condition presents an imminent danger of death or serious injury; and
   • The urgency does not allow sufficient time to eliminate the danger without the refusal.
3. Notification (where possible):
   The employee has notified the carrier of the hazardous condition and the intention not to perform further work unless the condition is immediately corrected.

Although the Tenth Circuit did not parse every prong in detail, it held that Berberich failed fundamentally on the two bedrock predicates:

• There was no “refusal to work” at all; and
• There was no “imminent danger” or existing hazardous condition at the time he acted.

V. Detailed Analysis

A. The Court’s Interpretation of “Refusing to Work”

1. Plain‑meaning analysis

The court applied standard textualist methodology, citing Smith v. Board of Governors of the Federal Reserve System, 73 F.4th 815 (10th Cir. 2023), and Robinson v. Shell Oil Co., 519 U.S. 337 (1997):

• Start with the statutory language;
• Consider the specific and broader context; and
• If the text is unambiguous, its plain meaning controls.

Applying this, the court reasoned that:

• To “refuse to work” ordinarily means to decline to perform work—i.e., non‑performance or cessation of assigned duties.
• Berberich, by his own repeated testimony, was doing exactly the opposite: he was performing his assigned work as a conductor.

He testified that:

• “[Conductor]s switch cars, pull pans, line switches, and I did what I … was supposed to”;
• Throwing the switch was “my job.”

Thus, the panel concluded it “cannot plausibly be argued” that acting in accordance with one’s normal duties constitutes “refusing to work” under § 20109(b)(1)(B).

2. Rejecting the plaintiff’s broader construction

The plaintiff’s theory effectively tried to recast “refusal to work” as “refusal to work in the unsafe manner preferred or required by the employer.” Under this reading, performing one’s job safely in defiance of an alleged unsafe “standing order” would count as a protected refusal to work.

The court rejected that construction as both:

• Factually unsupported (there was no concrete evidence of any order directed at him at the time to have the engineer line the switch); and
• Legally unsound in view of the FRSA’s structure.

Structurally, Congress had already, in § 20109(a)(2), explicitly protected employees for “refus[ing] to violate… Federal law, rule, or regulation relating to railroad safety.” If Congress had intended § 20109(b)(1)(B) to protect employees for refusing to violate internal policies or informal orders, it could have done so using similar language. It did not.

Instead, § 20109(b) is narrower: it focuses on situations where the employee stops working in the face of an existing hazardous condition posing an imminent threat. The court therefore held that expanding “refusing to work” to include “refusing to work in an unsafe way” would amount to an improper judicial amendment of the statute.

B. The Absence of a “Hazardous Safety Condition” and “Imminent Danger”

1. Statutory requirement of imminence

Section 20109(b)(2)(B)(i) requires that a reasonable person would conclude that the hazardous condition presents an “imminent danger of death or serious injury.” The court drew on Black’s Law Dictionary (12th ed. 2024) to define “imminent” as “threatening to occur immediately; dangerously impending.”

Under this standard:

• The danger must be concrete and about to occur—not speculative or contingent on a chain of events that might or might not happen.
• The situation must be such that, absent refusal, death or serious injury is about to occur or is highly likely in the very near term.

2. No evidence of any actual or imminent hazard at the time

The only “hazard” posited by Berberich was the risk that an engineer might leave the controls unattended to line the switch, with the train parked solely on its air brakes. But the court emphasized:

• There was no evidence that, on the specific January 2019 occasion, the engineer had been ordered to line the switch;
• There was no evidence the engineer intended to leave the controls or believed he should do so;
• By Berberich’s own deposition, the engineer merely:
  • Reported that the switch was aligned against the train; and
  • Expected the conductor (Berberich) to come line it, in keeping with “common practice.”

Thus, at the crucial moment:

• The engineer was not in the process of performing any dangerous act.
• No immediate plan existed for the engineer to abandon the controls to line the switch.
• What Berberich feared was at most a generalized possibility that supervisors, including Pollard, might at some point demand that an engineer line a switch in order to expedite a move.

Such speculative, contingent concerns do not satisfy the statute’s requirement of an imminent hazard “threatening to occur immediately.” Because the precondition of a qualifying hazardous condition was not met, there could be no protected refusal to work under § 20109(b)(1)(B).

C. The “Standing Order” Problem

The Tenth Circuit also underscored an evidentiary point that has both substantive and procedural significance: the label “standing order” was not supported by the record.

• On cross‑examination, Berberich conceded he was not aware of any written directive, bulletin, or formal rule requiring engineers to line switches generally.
• He described only “a lot of radio chatter” where Pollard, in particular circumstances, told engineers to get off and line a switch “to expedite movements.”
• Spigarelli’s testimony likewise described specific radio conversations with trains, not a blanket, yard‑wide mandate.

The court therefore characterized these as occasional, situation‑specific orders or preferences, not a “standing order” in the sense of a fixed, always‑applicable rule. That distinction matters in FRSA analysis because:

• If there is no generally applicable instruction, the claim that the employee is “refusing” to comply with a dangerous practice becomes weaker, especially absent a specific directive on the day in question.
• It also undermines any attempt to show that there was a consistent employer policy that could be characterized as an “unsafe order” to which Berberich’s actions were an exception.

D. Relationship to Other FRSA Protections

The decision implicitly clarifies the boundaries between different FRSA protected‑activity clauses:

• § 20109(a)(2) is directed at refusals to violate federal safety laws or regulations. If an employer orders a worker to do something that would violate a FRA regulation or other federal safety rule, the employee’s refusal is protected.
• § 20109(b)(1)(B), by contrast, focuses on refusal to continue working in the face of an existing imminent hazard, regardless of whether any underlying federal standard has been violated.

Berberich’s theory blurred this distinction by attempting to treat a perceived unsafe practice (allowing an engineer to line a switch while leaving controls unattended) as either:

• A de facto violation of safety rules (KCSR’s internal rule about not relying solely on air brakes); or
• A hazardous condition that he “refused” to participate in by doing the task himself instead.

The Tenth Circuit’s refusal to extend § 20109(b)(1)(B) to this scenario underscores that:

• Internal safety rules alone do not automatically transform an employer’s preference into a federally protected refusal under § 20109(b)(1)(B); and
• To invoke § 20109(b)(1)(B), there must be both:
  • A genuine, imminent hazard confronting the employee; and
  • An actual decision not to perform work because of that hazard.

E. Summary Judgment and the “Complete Failure of Proof” Doctrine

The court relied on standard summary‑judgment principles, most prominently Celotex Corp. v. Catrett and its own decision in Rocky Mountain Prestress, LLC v. Liberty Mutual Fire Ins., 960 F.3d 1255 (10th Cir. 2020):

• Summary judgment is appropriate when there is no genuine dispute of material fact and the movant is entitled to judgment as a matter of law.
• There must be more than a “mere scintilla” of evidence; the evidence must allow a reasonable jury to find for the non‑moving party.
• A “complete failure of proof concerning an essential element” makes all other facts immaterial.

Because the court found a complete absence of evidence supporting a protected “refusal to work” or the presence of a qualifying hazardous condition, it deemed other debates—such as whether Devenney knew of the incident, or whether Pollard’s animus could be imputed via a “cat’s paw” theory—immaterial.

F. Appellate Practice: Record Citation and Waiver

The opinion also reinforces an important appellate practice principle: appellate courts do not scour the record for uncited evidence or build arguments the parties have not raised. Citing Cordova v. Aragon, 569 F.3d 1183 (10th Cir. 2009), and Gross v. Burggraf Construction Co., 53 F.3d 1531 (10th Cir. 1995), the panel declined to consider potentially helpful testimony that was:

• In the record; but
• Not cited in Berberich’s appellate briefs.

Example: On redirect before the ALJ, Berberich’s counsel restated the alleged practice as a “preference” by managers that engineers throw switches “when it would expedite a move.” The panel noted that this testimony:

• Was not cited on appeal; and
• In any event, merely confirmed that the practice was situational (“when it would expedite a move”), not a blanket “standing order.”

For practitioners, the message is clear: in FRSA appeals (as in other cases), every piece of evidence supporting one’s theory must be expressly cited and integrated into the legal argument; otherwise, it may be treated as effectively waived.

G. The “Cat’s Paw” Theory Left Unresolved

Berberich urged a “cat’s paw” theory, invoking EEOC v. BCI Coca‑Cola Bottling Co. of Los Angeles, 450 F.3d 476 (10th Cir. 2006). In such cases:

• A biased subordinate without formal decision‑making authority manipulates the actual decisionmaker into taking an adverse action (e.g., termination).
• The decisionmaker becomes a “cat’s paw” who unwittingly carries out the subordinate’s discriminatory or retaliatory intent.

Here, the theory would have been that Pollard, incensed by the January 2019 incident, influenced Devenney to terminate Berberich for a pretextual safety violation (leaning on a railcar). However, the court explicitly declined to reach this issue because the lack of protected activity eliminated the FRSA claim at the threshold.

The opinion therefore does not limit the availability of cat’s paw theories in FRSA cases. Rather, it affirms a more basic point: regardless of motive, there can be no FRSA retaliation where the employee never engaged in protected activity in the first place.

VI. Precedents Cited and Their Influence

A. Lincoln v. BNSF Railway Co., 900 F.3d 1166 (10th Cir. 2018)

Lincoln is the Tenth Circuit’s leading FRSA decision and provided the elemental framework here:

• It set out the four‑part prima facie burden for FRSA retaliation claims.
• It emphasized that an employee must first show he engaged in protected activity before courts even reach knowledge, adverse action, or causation questions.

In Berberich, the panel repeatedly cited Lincoln as authority for:

• The prima facie structure; and
• The rule that a claim fails when the plaintiff cannot establish the protected‑activity element.

B. Wise v. DeJoy, 71 F.4th 744 (10th Cir. 2023)

Wise stands for the proposition that an appellate court may affirm a judgment on any ground supported by the record, even if that ground was not the basis of the district court’s decision, provided the appellant had a fair opportunity to address it.

Invoking Wise, the Tenth Circuit exercised its discretion to affirm on the alternative ground that Berberich lacked evidence of protected activity, rather than reviewing the district court’s knowledge analysis. This underscores that appellate courts may—and in FRSA cases often will—focus on whichever element presents the clearest dispositive issue.

C. Rocky Mountain Prestress and Celotex

Rocky Mountain Prestress, LLC v. Liberty Mut. Fire Ins., 960 F.3d 1255 (10th Cir. 2020), and Celotex Corp. v. Catrett provided the underlying summary‑judgment framework:

• Rocky Mountain articulates what counts as a “genuine dispute” and “more than a mere scintilla” of evidence.
• Celotex underscores that the absence of proof on an essential element renders all other factual disputes immaterial, justifying summary judgment.

Applied to FRSA, this means that once the court concluded there was no evidence of a protected refusal to work or an imminent hazard, disputes about Devenney’s knowledge, pretext, or the severity of discipline became irrelevant.

D. Smith, Robinson, and Statutory Interpretation

Smith v. Board of Governors and Robinson v. Shell Oil provided the interpretive tools for construing “refusing to work” and “imminent danger”:

• Both endorse a contextual plain‑meaning approach—examining the term itself, its immediate context, and its place in the broader statutory scheme.
• In Berberich, this led the court to emphasize the difference between § 20109(a)(2) (refusing to violate federal law) and § 20109(b)(1)(B) (refusing to work in hazardous conditions).

E. Cordova, Gross, and Record Use

Cordova v. Aragon and Gross v. Burggraf Construction Co. were cited to justify the court’s refusal to consider uncited record material. They reflect the Tenth Circuit’s consistent position that:

• Appellate courts do not act as “archaeologists,” digging up uncited evidence to potentially reverse or alter the outcome; and
• Counsels’ record citations and arguments determine the universe of evidence for appellate review.

F. EEOC v. BCI Coca‑Cola and the Cat’s Paw Concept

While the court referenced BCI Coca‑Cola to describe cat’s paw liability, it did not interpret or limit that doctrine in the FRSA context. Instead, the citation primarily served to frame what it was declining to decide.

VII. Complex Concepts Simplified

A. “Protected Activity” Under FRSA

“Protected activity” is conduct that Congress has chosen to shield from employer retaliation. Under FRSA, this includes, among other things:

• Reporting safety concerns or violations;
• Refusing to violate federal safety laws or regulations;
• Refusing to work in the face of imminent hazardous conditions (subject to strict conditions);
• Reporting workplace injuries; and
• Participating in investigations or proceedings related to safety complaints.

Not every disagreement over safety practices or every cautious decision by an employee is automatically protected. The conduct must fit within one of the enumerated categories and meet any statutory conditions (such as “imminence” and “no reasonable alternative” for refusals to work).

B. “Refusal to Work” Versus “Working Safely”

Under § 20109(b)(1)(B), “refusing to work” means:

• Stopping or declining to perform assigned work because of a hazardous condition;
• As opposed to simply choosing a different, safer way to perform the same work when no one is actively ordering unsafe conduct in that moment.

In Berberich, walking to throw the switch himself and then returning to other duties was characterized as “working,” not “refusing to work,” even if management might have preferred a faster but allegedly riskier method.

C. “Imminent Danger”

A danger is “imminent” when:

• It is about to happen; and
• There is a serious, immediate risk of death or serious injury.

Hypothetical or eventual dangers—things that could become dangerous under different circumstances—do not qualify. In this case, the fact that engineers might sometimes be asked to leave the controls did not constitute an “imminent danger” in a situation where the engineer was simply sitting in the cab and reporting a misaligned switch.

D. “Cat’s Paw” Theory

The cat’s paw theory comes from a fable: a monkey convinces a cat to pull chestnuts from a fire; the cat gets burned while the monkey enjoys the chestnuts. In employment law, it describes situations where:

• A biased supervisor or subordinate (the “monkey”) manipulates an otherwise neutral decisionmaker (the “cat”) into firing or disciplining someone for unlawful reasons.
• The decisionmaker may not personally harbor discriminatory or retaliatory animus but relies heavily on biased information when making the adverse decision.

In FRSA cases, cat’s paw can matter when a safety‑hostile manager engineers an investigation and the ultimate decisionmaker rubber‑stamps a recommendation without independent review. Berberich did not resolve any new issues about cat’s paw; it simply never got that far.

E. “Trains on the Law” and the Hours of Service Act

The ALJ, quoted by the Tenth Circuit, explained that “trains on the law” refers to trains whose crews reach the 12‑hour cap on consecutive work hours under the Hours of Service Act, 49 U.S.C. § 21103(a). That phrase appeared in testimony where Devenney allegedly complained about being “tired of hearing [Berberich’s] name come up on trains on the law,” suggesting repeated delays or long duty tours.

That context might have supported a narrative of management frustration with safety‑related delays. But under Berberich, unless such conduct fits one of the enumerated “protected activities,” it cannot form the basis for FRSA retaliation liability.

VIII. Impact and Significance

A. Narrowing of FRSA § 20109(b)(1)(B) in the Tenth Circuit

Berberich is, in practical terms, a narrow but important decision:

• It is one of the first published Tenth Circuit cases to interpret in detail what constitutes a “refusal to work” and a qualifying hazardous condition under § 20109(b)(1)(B).
• It makes clear that:
  • Employees must show actual non‑performance of work, not merely different or cautious performance;
  • The danger must be present and imminent, not hypothetical or historical; and
  • General workplace safety disagreements are not, by themselves, protected refusals.

For future FRSA plaintiffs in the Tenth Circuit, this raises the bar for using § 20109(b)(1)(B) as a retaliation platform. Claims framed as “I performed the job more safely than my supervisor preferred” will face significant headwinds unless they involve:

• A clear, contemporaneous order to perform work under unsafe, imminently dangerous conditions; and
• A clear refusal to do the work (or to continue working) because of that danger.

B. Strategic Implications for Railroad Employees and Counsel

For employees and their attorneys, Berberich suggests several strategic points:

• Where the core complaint is being told to violate internal rules or perform work unsafely in a way that would violate federal law or regulation, § 20109(a)(2) may be a stronger basis than § 20109(b)(1)(B).
• To invoke § 20109(b)(1)(B), counsel should:
  • Document the specific directive or working condition alleged to be imminently hazardous;
  • Clarify that the employee actually declined or stopped work (e.g., refused to move the train, refused to go into a confined space, refused to climb unsafe equipment); and
  • Preserve and cite all supporting testimony and evidence in the record and on appeal.

C. Employer Compliance and Risk Management

For rail carriers, the case provides both reassurance and caution:

• Reassurance:
  Employers are not automatically exposed to FRSA liability every time an employee chooses to perform work in a slower, more cautious way, absent a genuine refusal to perform assigned tasks in the face of immediate danger.
• Caution:
  The opinion does not diminish protections where employees:
  • Refuse to violate federal safety rules; or
  • Truly refuse to work in an imminent, dangerous condition (for example, operating equipment known to be defective in a way that could immediately kill or seriously injure).

Moreover, Berberich does not insulate employers from retaliation claims premised on other forms of protected activity (e.g., injury reporting, participation in safety investigations, or refusals under § 20109(a)(2)).

D. Administrative Proceedings vs. De Novo District Court Review

The case also illustrates the interplay between:

• OSHA and ALJ determinations under the Department of Labor’s administrative process; and
• De novo review in federal district court under § 20109(d)(3) when no final agency decision issues within 210 days.

While the district court and Tenth Circuit were not bound by the ALJ’s findings, several of the ALJ’s conclusions (e.g., that Berberich did not “refuse to work,” that there was no “standing order,” and that supervisors only occasionally had engineers throw switches) were echoed in the appellate opinion. This reflects how, in practice, administrative fact‑finding can shape the evidentiary record and narrative even though district‑court review is technically de novo.

E. Possible Unresolved Questions

The opinion leaves several issues open for future cases:

• How far the concept of “refusal to work” might extend in borderline situations where an employee suspends some duties while continuing others;
• How courts will treat cases where a supervisor explicitly orders unsafe conduct at the moment and the employee simultaneously performs some tasks while refusing others;
• The precise interaction between § 20109(a)(2) and § 20109(b)(1)(B) when both might arguably apply (e.g., where an unsafe order would directly violate a federal regulation and also creates an imminent hazard).

IX. Conclusion

Berberich v. Kansas City Southern Railway Company sets an important precedent in the Tenth Circuit on the meaning of “refusing to work” and “hazardous safety condition” under FRSA § 20109(b)(1)(B). The court held that:

• Merely performing one’s job in a careful, arguably safer way—here, a conductor walking to line a switch himself—is not a “refusal to work” within the statute’s plain meaning.
• A theoretical or generalized risk that management may sometimes favor unsafe practices does not constitute an “imminent danger” or present hazardous condition absent concrete evidence of a dangerous act about to occur in the particular situation.
• Without a genuine refusal in the face of a concrete, imminent hazard, there is no protected activity under § 20109(b)(1)(B), and FRSA retaliation claims fail at the threshold.

By grounding its reasoning in the statutory text, structure, and well‑established summary‑judgment and appellate‑practice doctrines, the Tenth Circuit provided both guidance and limits for future FRSA litigants. Employees, unions, and railroads must now navigate a clarified landscape in which:

• Plain‑meaning statutory interpretation will drive outcomes; and
• Careful factual development and precise characterization of alleged “refusals” and “hazards” will often be dispositive.

In sum, Berberich stands for the principle that FRSA’s “refusal to work” protections are potent but carefully cabined: they apply when an employee truly stands down in the face of an imminent, life‑threatening danger—not when the employee is simply doing the job more safely than a supervisor might prefer.