Workers believe the best or healthiest people leave first when a toxic coworker goes unaddressed: 44% say so in GCheck’s The Consequence Gap Report. Their own records add detail. In past cases the toxic coworker more often left first (57%), and workers left while the coworker stayed at 40% where the coworker was protected, against 10% where removed.
Key Takeaways
- GCheck’s survey of 1,500 workers found that 44% believe the best or healthiest people leave first when a toxic coworker goes unaddressed.
- Workers’ own records add detail: among the 630 workers whose exposure is behind them, the toxic coworker more often left first (57%).
- Departures track the organization’s response. Workers left while the toxic coworker stayed at 40% where the coworker was rewarded or protected, 24% where nothing happened, and 10% where the coworker was removed, a documented pattern.
- The belief and the record describe different slices of experience, so both can hold: the belief speaks to situations that go unaddressed, and the record covers cases that have ended.
- Exit data is more informative when it is read against a written definition of “best,” compared across teams, and paired with other sources, since some departing employees aren’t candid and some leave before an exit interview happens.
- An organization can check the belief in its own data by comparing leavers and stayers on the same team around the response to a concern, using one definition applied to everyone.
What Workers Believe, and What Their Own Record Shows
The Belief and the Record
The survey asked workers about leaving in two different ways. The first was a belief: 44% say the best or healthiest people leave first when a toxic coworker goes unaddressed. The second was a record of what happened to the workers themselves:
| What the survey asked | Figure | Kind of evidence |
| Belief: the best or healthiest people leave first when a toxic coworker goes unaddressed | 44% | A stated belief (PERCEPTION) |
| Record: among workers whose exposure to a toxic coworker is behind them (630 workers), the toxic coworker left first | 57% | Workers’ own accounts (INCIDENCE) |
| Pattern: workers left the job while the toxic coworker stayed, by what the organization did (283, 477, and 476 workers) | 40%, 24%, 10% | Workers’ own accounts (INCIDENCE) |
The report’s reading is that both can hold at once: in cases that ended, the coworker more often went first, and where coworkers are protected, departures shift toward other people.
The 630 Past Cases
The 57% describes cases that have ended. Of the 1,272 workers who have had a toxic coworker, 630 are in the group whose exposure is behind them. The remaining 642 are outside that group, and the report’s split between past and current exposure implies those situations are still in progress.
That makes the record a worker looking back on a case that is over and saying who went first, a clear and countable account of what happened.
The survey puts the worker’s own account at the center. Its findings come from workers describing themselves and what they saw, in a survey of 1,500 US adults employed full time in August 2026, weighted to the census on age, gender, and region.
Reading the Belief and the Record Together
Two Lenses on the Same Question
The belief and the record look at the same situation from different angles. The belief is stated in terms of the best or healthiest people, a category people define in different ways: strong results, strong teamwork, strong attendance, or simply someone they like working with. The record is stated in terms of order of departure, which can be counted. Each contributes something. The belief shows how workers read these situations, and the record shows what happened to them. Neither is the more truthful of the two, because they answer different questions.
That is why both can hold at once. The belief speaks to situations that go unaddressed, and the record covers cases that have ended. A worker can watch a protected coworker stay while teammates leave, and another worker can look back on a case that closed with the coworker’s exit, and each account is accurate.
The report tracks who left, in what order, and what the organization did about the coworker. How strong each person’s performance is recorded in an organization’s own review systems, which is why the later sections show how to bring the two together.
Two Figures About Toxic Coworkers Leaving
Two figures from the research describe different things. Among the 1,272 workers who have had a toxic coworker, nothing happened to the coworker in 38% of cases, the coworker was rewarded, promoted, or protected in 23%, and only 36% eventually left or were removed. The 57% above answers a different question: in cases that had ended, who went first.
One figure describes the coworker’s fate across all exposed cases, including ongoing ones. The other describes the order of departure in ended cases. Each is accurate on its own terms.
Keeping them apart matters in practice. Quoting the 57% alone can suggest that toxic coworkers reliably go, while quoting the 36% alone can suggest they rarely do.
What the Data Shows About Who Leaves
Departures Track the Organization’s Response
The clearest finding about leaving compares workers by what the organization did about the toxic coworker. It asked who left first, the worker or the coworker, and reports the share of workers who left the job while the coworker stayed:
| What happened to the toxic coworker | Share of workers who left while the coworker stayed |
| Rewarded or protected (283 workers) | 40% |
| Nothing happened (477 workers) | 24% |
| Removed (476 workers) | 10% |
The comparison is a pattern: the groups differ, each base includes workers who are still exposed and so haven’t left, and the figures show where departures cluster. Workers left first four times as often where the coworker was rewarded or protected as where the organization acted, and where nothing happened, 24% left first, more than twice the 10% where the coworker was removed. For an HR team, the grouping suggests a question to put to its own exits: what had the organization done about the coworker by the time each person left?
Order and Reason
The table records who left while the coworker stayed. Reasons come from the workers themselves. One worker, describing a case where the coworker was later removed, put it this way: “I did not leave because of them, but their constant negativity and undermining made the workplace increasingly stressful and exhausting.”
The account is a single voice and isn’t evidence for any percentage. It shows how departure and reason can come apart: a worker can leave, or stay, for reasons the order of departure doesn’t record. That is why the next section treats exit data as something to question, not just count.
For an organization, the practical point is that a departure date records timing, and a reason requires asking, ideally more than once and from more than one direction: the departing employee, the people closest to them, and the pattern across the team.
How to Read Your Own Exit Data
Six Practices for Testing the Belief
An organization’s own exit data is better evidence about its own workplace than any survey, provided it is read carefully. Six practices keep the analysis honest:
- Separate three questions: who left, in what order, and why. A count of leavers answers none of them, and order needs the date each person left relative to the person of concern.
- Define “best” before counting, in writing. Use one criterion, such as performance ratings or a regretted-loss designation, and apply it to every departure. Any definition carries its own biases, so review it for consistency and don’t adjust it after seeing who left.
- Compare leavers with stayers on the same team around the time of a concern and the response to it, not against company-wide turnover rates, which blur the teams where something is going on.
- Group departures by what the organization did: rewarded or protected, nothing, or acted. This is the survey’s own grouping applied internally, and it is the comparison where the clearest pattern appeared.
- Keep open cases in the count. Situations still in progress belong in the data, or the analysis will describe only the ones that have ended.
- Treat small groups as directional. A team with three departures can suggest a question and can’t support a firm conclusion, and groups that small shouldn’t be reported in a way that identifies anyone.
Taken together, these turn “the good people are leaving” from an impression into a question with an answer. The answer may be yes, no, or different by team, and any of the three is more useful than the belief.
What Exit Interviews Add, and What Else Helps
An article by Anthony Klotz and Mark Bolino, published on SHRM’s site as a reprint from Harvard Business Review, makes the case for supplementing exit interviews. Some employees leave before one can happen, including those who stop showing up or quit abruptly. Research the authors cite suggests that a large percentage of employees who do take part are not candid. Some soften critical feedback to leave a good impression, some see it as a waste of time because they expect no change, and some feel management has not earned an honest answer.
One worker’s sentence shows how that plays out: “Got tired of the negativity and needed a new job.” An exit form could record only the second half, which is the stated reason. The first half is closer to the cause. The authors suggest three additions, each of which supplies a view that a departing employee’s own words don’t:
- Code how people resigned and review the pattern. The piece describes seven resignation styles, from a grateful goodbye to bridge-burning to impulsive quitting, and suggests that clusters in one department or under one supervisor are worth investigating.
- Speak, voluntarily, with the coworkers closest to the person who left. The authors acknowledge that this can feel like asking people to be disloyal to a friend, so participation should be optional and its purpose, improving the experience of the remaining workforce, stated plainly.
- Follow up on where alumni go, using information they choose to share. A pattern of exits to one competitor or one kind of role says something that an exit form can’t.
Cautions When Collecting It
If an exit interview or a conversation with coworkers surfaces an allegation of harassment, discrimination, or retaliation, a safety problem, or a medical or accommodation issue, it goes to HR or counsel. According to the EEOC, an employer can be liable for harassment by non-supervisory employees if it knew or should have known about it and failed to take prompt and appropriate corrective action, so what an exit conversation surfaces can carry obligations of its own, and a different standard can apply when the person accused is a supervisor.
When coworkers are asked about a departure, the questions should seek their own view and stop short of probing who said what to whom or asking for medical details. The National Labor Relations Board states that employees, in union and non-union workplaces alike, have the right to talk with coworkers about wages and working conditions, and that an employer cannot discipline or threaten them for it or coercively question them about it. Individual comments also belong in confidence: analyzed in aggregate, handled the same way for every departure, and never reported in groups small enough to identify anyone.
This is a practice note, not legal advice. Employment and privacy rules vary by jurisdiction and by the facts of a case, and an organization that is unsure how to handle what it has heard should ask before acting.
Where Hiring Fits
A background check does necessary, effective work at the point of hire, applied within the federal, state, and local rules that govern what can be considered, when, and how. It confirms what is on record. A pattern in exit data is a different kind of signal: clusters of departures around one person, or around one response, can be red flags about what happens on the job after a hire, which no record at hire can show.
A pattern is a reason to look, not a finding about any person, and any review that follows should give the person involved the same process as anyone else. Where it does lead to the hiring and promotion history, the history is worth revisiting, and for future hires into similar roles a conduct-focused reference conversation, held with the candidate’s knowledge and consent, asks the people who worked alongside a candidate. GCheck’s guide to reference check questions covers that ground. When a third-party provider assembles a reference report, the Fair Credit Reporting Act‘s disclosure and written-authorization requirements can apply.
Exit patterns can also feed forward into onboarding. Where a team has lost people around one situation, the next hire into that team deserves a clear description of conduct expectations and a named person to raise a concern with, so the new arrival isn’t left to learn the unwritten rules from the people who stayed.
What This Isn’t Arguing
Nothing here says the belief is wrong. In some workplaces the best people may well leave first, and an organization’s own performance ratings are where that can be seen. The research contributes to the record of who went first and what the organization did about the coworker.
It also isn’t a case for ranking workers or for ending exit interviews. A definition of “best” is useful for one narrow purpose, checking the belief, and it becomes a problem the moment it turns into a label that decides who gets attention. Exit interviews keep their value as one source among several.
Treating the belief as settled carries its own risk. A manager who assumes the best people are the ones leaving may reach first for retention offers and stay-bonus conversations, and in doing so skip the conduct question that may be driving departures on the team.
What This Means for Compliance
Transparent Compliance, applied to exit data, means being clear about what the numbers show: defining terms before counting, documenting departures the same way each time, and telling a team in general terms that the pattern is being reviewed. A belief is a starting point, and a record is what lets an organization check it.
Fair Compliance supplies the other half: the same definition and the same examination for every departure, whoever the leaver is. If a departure draws an investigation because the person was seen as a star and another gets none because the person was seen as ordinary, the standard has a performance threshold in it, and the exit data can reflect that bias back at the organization.
A sensible place to start is one team and its last year of departures. List each leaver, the date, and what the organization had done about any open concern, keeping the list in confidence, then check that the “best” definition was written down before the list was read. The result won’t settle the belief for the whole company. It will show whether the question deserves a larger look.
Frequently Asked Questions
Do good employees quit first when a toxic coworker stays?
Workers believe so: GCheck’s The Consequence Gap Report found that 44% say the best or healthiest people leave first when a toxic coworker goes unaddressed. Workers’ own records add detail. In past cases the toxic coworker more often left first (57%), and 40% of workers left while the coworker stayed where the coworker was rewarded or protected, against 10% where removed.
What percentage of workers believe good employees leave first?
In GCheck’s survey of 1,500 workers, 44% said the best or healthiest people leave first when a toxic coworker goes unaddressed. It is a stated belief, and the survey pairs it with workers’ own records: among past cases, the toxic coworker more often left first (57%).
Who leaves first, the worker or the toxic coworker?
Among the 630 workers in GCheck’s research whose exposure to a toxic coworker is behind them, the coworker more often left first (57%). That covers ended cases only. Where the organization rewarded or protected the coworker, 40% of workers left while the coworker stayed, compared with 10% where the coworker was removed.
Do exit interviews show why good employees leave?
They show part of the picture. An article on SHRM’s site, reprinted from Harvard Business Review, notes that research suggests a large percentage of departing employees are not candid, and that some leave before an exit interview can happen. Pairing exit interviews with other sources, such as the views of the coworkers closest to the person who left, fills in the rest.
How can I test whether good employees are leaving first in my organization?
Write down a single definition of “best” before counting. Then compare leavers with stayers on the same team around the time of a concern, group departures by what the organization did, keep open cases in the count, and treat small groups as directional. Apply the same method to every departure and handle individual exit comments in confidence.
ABOUT THE CREATOR
Charm Paz, CHRP
Recruiter & Editor
Charm Paz is an HR professional at GCheck, specializing in background screening, fair hiring, and regulatory compliance. She holds from the Professional Background Screening Association (PBSA) and helps organizations navigate employment regulations with clarity and confidence.
With a background in Industrial and Organizational Psychology, she translates policy into practice to build ethical, compliant, human-centered hiring systems that strengthen decision-making over time.
