Workplace Bullying Investigation: 7 Controls That Protect People and Evidence
Workplace bullying investigations fail when they treat harm as an HR dispute while the work system that allowed the behavior remains untouched.
Principais conclusões
- 01Workplace bullying should be investigated as a psychosocial risk because the work system may allow the exposure to continue.
- 02The first control is stabilization, especially when the person accused controls schedules, performance ratings or task allocation.
- 03A strong investigation translates broad labels into observable conduct, work context, evidence and consequences.
- 04Confidentiality means controlled disclosure, not an impossible promise of total secrecy.
- 05Corrective action should address individual conduct, leadership control and the work-system conditions that allowed the behavior.
Workplace bullying is often investigated as if it were a private conflict between two people. That lens is too narrow for occupational health and safety. When intimidation, humiliation, isolation or repeated hostile behavior happens inside a work process, the company must ask a harder question: what conditions allowed the exposure to continue?
This guide is written for EHS managers, HR leaders and operations directors who need a disciplined investigation process. The thesis is direct: a workplace bullying investigation only protects people when it preserves evidence, controls retaliation and changes the work conditions that allowed the behavior to become normal.
Why workplace bullying is a psychosocial risk, not only an employee relations case
ILO Violence and Harassment Convention, 2019 (No. 190) treats violence and harassment in the world of work as conduct, practices or threats that can result in physical, psychological, sexual or economic harm. That definition matters because it moves the issue beyond personality conflict and places it inside the employer's duty to prevent harm.
ISO 45003:2021 also places psychosocial risk inside occupational health and safety management. The standard does not turn HR into EHS, and it does not turn EHS into HR. It requires both functions, with operations, to examine how work is organized, supervised and corrected when exposure appears.
Across 25+ years in executive EHS and safety culture work, Andreza Araujo has seen that organizations often discover bullying too late because the first signs were treated as interpersonal friction. In Safety Culture: From Theory to Practice, Araujo describes culture as repeated leadership behavior. That point is decisive here because bullying survives when repeated behavior is tolerated by supervisors, peers or performance pressure.
1. Stabilize the work situation before collecting the full story
The first control is stabilization. Before the organization interviews ten people or requests written statements, it must decide whether the reported exposure can continue during the investigation. If the same supervisor controls schedules, overtime, performance ratings or task allocation for the complainant, the company may be collecting testimony while the risk remains active.
Stabilization can include temporary reporting-line changes, schedule separation, remote work where feasible, a neutral contact person, limits on direct communication and documented instructions against retaliation. The measure should protect the process without punishing either party before facts are established.
The common error is to call for calm and leave the work structure untouched. That approach looks neutral, although it may preserve the exact power arrangement that made reporting difficult. An investigation that cannot protect the person who speaks will teach the rest of the workforce to stay silent.
2. Define the allegation in observable work terms
Bullying allegations become hard to investigate when the company records only broad labels such as toxic behavior, harassment or disrespect. Those words may be true, but they are not enough for evidence, control selection or management accountability. The allegation has to be translated into observable work terms.
A useful investigation question is not only, what happened? It is also, where did it happen, who was present, what work decision was affected, what changed afterward and what pattern existed before the formal complaint? This shifts the inquiry from emotion alone to conduct, context and consequences.
Examples include repeated public humiliation in daily meetings, exclusion from essential task information, impossible deadlines assigned selectively, threat of shift changes after a safety concern, denial of breaks, hostile messages outside working hours or mocking an employee after an error. These details allow HR and EHS to separate isolated conflict from recurring exposure.
3. Preserve evidence before memory and records decay
Evidence in bullying cases is often fragile because many behaviors occur verbally, informally or through repeated small acts. That does not make the case impossible. It means the organization must preserve available evidence quickly and consistently.
Relevant evidence may include messages, meeting invitations, task assignments, schedule changes, overtime records, performance notes, transfer requests, absence patterns, complaint logs, previous informal reports and witness accounts. The investigator should also check whether similar reports have appeared in climate surveys, exit interviews or speak-up channels.
This is where the existing article on speak-up metrics connects to bullying prevention. If the organization tracks only formal complaints, it misses weak signals. If it tracks recurring silence, withdrawal and repeated informal warnings, it can see the pattern before the case becomes severe.
4. Separate confidentiality from secrecy
Confidentiality protects people and evidence. Secrecy protects the organization from discomfort. The two are not the same, and confusing them damages trust.
The complainant should know who will receive information, why those people need it, what can and cannot be promised and how retaliation concerns will be handled. The respondent should receive enough information to answer the allegation fairly, without unnecessary exposure of private health details or unrelated personal history. Witnesses should understand the boundaries of their role and the expectation against retaliation.
A common trap is promising total confidentiality. In most investigations, that promise is impossible because facts must be checked with others. A better commitment is controlled disclosure, which means information is shared only with people who need it to investigate, protect or decide.
5. Interview for patterns, not only incidents
An isolated incident may violate policy, but bullying usually appears as a pattern. The interview plan should therefore test frequency, escalation, witnesses, changes in work allocation, prior complaints and the effect on participation, concentration, health, attendance and safety reporting.
Interviewers should avoid questions that sound like blame toward the person who reported. Questions such as why did you wait, why did you not confront the person or are you sure you did not misunderstand can shut down useful evidence. Better questions ask what made reporting difficult, what changed after each event and who had enough authority to intervene.
James Reason's work on latent failures is useful here because harmful events rarely depend only on the last visible action. A bullying case may reveal weak supervision, production pressure, unclear decision rights, fear of retaliation and a performance system that rewards abusive urgency. If the investigation stops at one person's behavior, it may miss the conditions that will reproduce the harm.
6. Decide corrective action at three levels
Corrective action should not be limited to discipline, even when discipline is necessary. A mature investigation separates three levels of response: individual conduct, leadership control and work-system exposure.
Individual conduct addresses confirmed behavior through discipline, coaching, role change or other measures consistent with policy and law. Leadership control addresses what the supervisor did or failed to do when signals appeared. Work-system exposure addresses workload, staffing, incentives, meeting routines, reporting channels, performance pressure and any condition that made abusive behavior easier to hide.
The article on psychosocial risk controls explains why this distinction matters. An organization can remove one aggressor and still keep the same work design that rewards fear, silence and impossible delivery. That is not prevention. It is replacement.
7. Close the case with follow-up, not only a letter
The formal closure of an investigation is not the same as risk closure. After the decision, the organization needs a follow-up rhythm to verify whether retaliation occurred, whether work arrangements are stable and whether the agreed controls are functioning under pressure.
Follow-up should include check-ins with affected parties, supervisor monitoring, review of schedule and task allocation, confirmation that witnesses were not punished informally and a review of weak signals in the team. If the case involved absence or treatment, the return-to-work process must respect privacy while still controlling the work exposure. The guide on return to work after mental-health absence is relevant when the investigation intersects with health-related absence.
Case closure should also feed the safety management system. If the same department, supervisor group or work rhythm appears repeatedly, the issue is not a collection of isolated complaints. It is a psychosocial exposure that leadership has to control.
Weak investigation versus protective investigation
The table below can help EHS, HR and operations test whether the process protects people or only protects the file.
| Investigation step | Weak approach | Protective approach |
|---|---|---|
| Initial response | Ask both sides to stay professional | Stabilize reporting lines, communication and retaliation controls |
| Allegation | Record broad labels | Translate the concern into observable conduct and work impact |
| Evidence | Rely mainly on memory | Preserve records, patterns, witnesses and weak signals |
| Confidentiality | Promise total secrecy | Explain controlled disclosure and process boundaries |
| Corrective action | Close with discipline or no discipline | Act on conduct, leadership control and work-system exposure |
| Closure | Send a final letter | Verify retaliation control, team climate and sustained risk reduction |
What EHS, HR and operations should do next
Start by reviewing the last twelve months of bullying, harassment, conflict, transfer, absence and exit-interview signals in one high-pressure area. Do not look only for formal complaints. Look for repeated names, repeated supervisors, repeated shift patterns, repeated urgency and repeated silence after someone raised a concern.
Then build one shared protocol between HR, EHS and operations. HR should own fairness, privacy and policy alignment. EHS should own psychosocial risk method, exposure controls and management-system learning. Operations should own work design, staffing, supervision and the correction of daily routines. Senior leadership should own the tradeoffs when controls conflict with production pressure.
Workplace bullying investigations are not a paperwork exercise. They are a test of whether the organization can hear harm early, protect people during the process and correct the work system that made the harm possible. If your leadership team needs to connect psychosocial risk, safety culture and investigation discipline, Andreza Araujo and ACS Global Ventures can support a practical diagnostic with EHS, HR and operations leaders.
Perguntas frequentes
Is workplace bullying a psychosocial risk?
Who should lead a workplace bullying investigation?
What evidence matters in a bullying investigation?
Can confidentiality be promised in a bullying complaint?
What should happen after the investigation closes?
Sobre a autora
Andreza Araujo
Global Safety Culture Specialist
Andreza Araujo is an international reference in EHS, safety culture and safe behavior, with 25+ years leading cultural transformation programs in multinational companies and impacting employees in more than 30 countries. Recognized as a LinkedIn Top Voice, she contributes to the public conversation on leadership, safety culture and prevention for a global professional audience. Civil engineer and occupational safety engineer from Unicamp, with a master's degree in Environmental Diplomacy from the University of Geneva. Author of 16 books on safety culture, leadership and SIF prevention, and host of the Headline Podcast.
- Civil Engineer (Unicamp)
- Occupational Safety Engineer (Unicamp)
- Master in Environmental Diplomacy (University of Geneva)