A company can have low overall turnover while one department keeps losing people through resignations and internal transfers. The starting point is not the company average. It is who left that department, when they left, and the conditions under which they worked. Rather than attributing everything to a hiring mismatch, examine unclear roles, an ever-growing workload, decisions left waiting, contradictory instructions, and what happens after employees raise concerns.
For example, even with annual company-wide turnover of 4%, every member of a small department’s original team could have resigned or transferred elsewhere within three years. There is no contradiction: annual company turnover and the proportion of a particular group still in its department three years later measure different populations over different periods.
The 4% figure, three-year period, and workplace situations in this article are hypothetical or generalized examples used to explain a structure. This is not an investigation verifying a particular company’s figures or an individual’s experiences. It also distinguishes findings from research from practical suggestions for examining a workplace.
1. Company turnover of 4% can coexist with zero retention of a department’s original team
Suppose a company has an average workforce of 500 and 20 people leave the company during a year. Using that denominator, annual turnover is 4%. Meanwhile, if all four people who belonged to one department at a particular point have left that department three years later, leaving only their replacements, retention of the original four within the department is 0%. That does not make the department’s annual turnover rate 100%. Internal transfers are not normally counted as departures from the company, either.
An unchanged headcount can conceal a loss of accumulated experience. Before comparing figures, align the periods, denominators, and definitions of resignation and transfer. Reorganizations, business closures, planned rotations, and promotions or transfers employees wanted can also replace a team. Replacement alone is not enough to conclude that a workplace is dysfunctional.
Useful breakdowns include department and direct supervisor, employment arrangement, new-graduate versus experienced hire, tenure, assigned work, and handover burden. Looking separately at retention after three months, six months, one year, two years, and three years helps distinguish initial adjustment problems from departures after employees have learned the job. Further breakdowns, such as age, should be used only where necessary and where individuals cannot be identified from small groups.
Record resignations, transfers out, leave of absence, long-term absence, transfer requests, and consultations separately. One person may seek help, take leave, and eventually resign. Adding these events into a single “exit rate” would count that person more than once. If a combined measure is needed, specify the observation period, population, deduplication rules, and what counts as an exit. Seeking advice must not itself become a marker of departure or of a “problem employee.”
2. Why a department can keep hiring while losing experience
When several people leave a small department, ordinary operations, unresolved cases, handovers, and newcomer training tend to concentrate on the experienced people who remain. Filling a vacancy does not mean the new employee can immediately work independently. Judging recovery by hiring numbers alone hides the support needed during that transition.
The losses extend beyond written procedures. A department can lose the reasons behind earlier decisions, contacts in other departments, actual deadlines, ways of handling exceptions, knowledge of previously failed approaches, and the experience needed to spot difficult cases early. If one person also interprets vague instructions for everyone else, their departure may reveal that they effectively were the department’s operating system.
Consider a team that expands, then loses several members. One remaining employee trains all the replacements and subsequently transfers out as well. Headcount recovers, but everyone left is relatively new. Breaking that cycle requires changes to work definitions, records, and workload distribution alongside recruitment. “We have newcomers, but people do not stay once they have developed” is a condition worth noticing.
3. Experienced hires still need an introduction to how this company works
An experienced professional does not automatically know a new employer’s decision-makers, responsible departments, exception procedures, or rules about whether a verbal instruction overrides a document. Saying only “You have experience, so you should be able to do this” asks them to use general professional experience to fill gaps in the company’s own arrangements.
At a minimum, explain:
- What the person is responsible for, and what falls outside that responsibility
- The purpose, priorities, completion criteria, and actual deadlines of the work
- What they can decide independently and what requires approval
- Where to ask questions, an alternative consultation route, and the final decision-maker
- Boundaries with other departments and how exceptions are handled
- Regular opportunities to check progress, and what existing work will be reduced when new work is assigned
Bauer and colleagues’ meta-analysis identifies role clarity, self-efficacy, and social acceptance as important elements in understanding newcomer adjustment. It also examines relationships with job satisfaction, organizational commitment, intention to remain, and turnover.[1] Research by Japan’s Institute for Labour Policy and Training examines relationships between support and onboarding measures for experienced hires and their organizational adjustment and intention to stay.[2] Neither body of research guarantees that introducing any single measure will stop departures.
As employees learn their jobs, they may notice duplicate data entry, rework after approval, tasks without an owner, or metrics that do not reflect reality. If such observations increased before someone left, investigate their substance before concluding that the person’s attitude deteriorated once they settled in. Experienced people leave for many reasons. It is no more appropriate to assume that everyone who leaves is highly capable than to assume that everyone who stays is compliant.
4. Are the people who notice gaps expected to fill all of them?
Some work happens before the visible task begins: clarifying its purpose, finding the necessary information, identifying stakeholders, adjusting deadlines, and deciding how to handle exceptions. If conditions that management should define remain unresolved, the employee takes on both execution and work design.
“Create training materials,” for example, is not a complete brief if the audience, scope, mandatory topics, reference date, final format, and approver are unknown. Clarification will continue after work starts. Add tracking sheets, recruitment initiatives, operating procedures, event planning, and everyday inquiries, and the burden becomes much greater than the visible deliverables suggest.
When conscientious employees fill these gaps every time, meeting the deadline does not mean the process has been fixed. If the company interprets success as proof that the current arrangement works, it remains dependent on the same people’s coordination and memory. Record the time spent filling gaps and specify who will decide what next time.
Adding work should also mean deciding what to stop, postpone, redistribute, or renegotiate. Adjustments to quality or scope must still meet legal, safety, and necessary quality standards. Keeping everything in place while saying “Find a way” may simply push the responsibility for setting priorities back onto the employee.
5. What separates delegation from handing down responsibility without authority?
Delegation includes sharing not only the intended result but also available resources, decision boundaries, and the circumstances in which consultation is needed. Asking an employee to develop the method is not inherently inappropriate. The aim is to combine discretion and support in a way that fits the work and the person’s experience.
Problems arise when a supervisor fixes the conclusion, rejects objections and deadline changes, and supplies neither the resources nor the decision criteria, while leaving the employee to work out implementation and explain any failure. If the employee raises a problem but the plan stays unchanged and is later judged against different standards, their decision rights do not match their responsibility.
An employee told to “think for yourself” can distinguish responsibility for proposing a method from responsibility for deciding its underlying conditions. For example: “I can develop options for the method. Before that, I’d like to confirm the budget limit, mandatory requirements, priorities among existing tasks, and final approver.” This makes the conditions for execution visible rather than leaving the discussion at simple disagreement.
6. Before turning an internal event’s benefits into “hours saved”
Imagine being asked, after a sports day or social event has been proposed, to show how many hours of work it will save. Numerical evaluation is not itself a problem. But assigning time savings to an initiative intended to encourage interaction, without supporting evidence, can turn evaluation into an exercise in producing a convenient reporting figure.
Start by asking:
- What problem is the initiative intended to address: interaction, ease of asking for help, or coordination between departments?
- What change would indicate progress toward that objective?
- What can be measured before and after, when, and under comparable conditions?
- How will busy periods, staffing changes, or other improvement initiatives be taken into account?
- How will the time and cost of preparation, participation, and evaluation be recorded?
If the aim is to make it easier to consult colleagues in other departments, perceptions of that ease and actual opportunities for contact can be examined. Participant satisfaction, however, is not the same as a reduction in work time. Even if a before-and-after figure changes, the event cannot be credited with the whole change unless the comparison supports that conclusion. Without a suitable comparison group or condition, state what was observed and the limits on causal inference.
The sequence should be problem, objective, expected change, indicator, measurement method, and comparison. Do not reverse it into a desired reporting figure, an existing event, and a calculation constructed to justify the figure. Managers must allow some effects to remain unmeasured and some estimates uncertain.
An event alone cannot solve everyday intimidation, excessive workload, or an uneven distribution of interruptions. Even where an interaction initiative has value, preparation and attendance should not create an additional burden. It needs to sit alongside improvements to ordinary working conditions.
7. Examine contradictory instructions through four separate lenses
“Take initiative, but don’t decide without permission.” “Ask if you don’t understand, but how can you not know that?” “Move quickly, but check everything first.” People sometimes call these situations a double bind. The label alone is not a psychological diagnosis or a legal finding. In practice, separating the problems makes them easier to address.
The first is role conflict: incompatible expectations. Examples include asking for improvements while prohibiting departures from precedent, or requiring coordination while forbidding contact with other departments. However, “Act independently within these boundaries and seek approval outside them” is not necessarily contradictory if those boundaries are consistent and clearly explained.
The second is role ambiguity: uncertainty about purpose, scope, completion criteria, priorities, or evaluation standards. Research has examined relationships between role conflict and ambiguity and job attitudes, strain, and other outcomes.[3] Tubre and Collins’ meta-analysis found a negative relationship between role ambiguity and job performance, but the relationship between role conflict and performance was small. The two should not be presented as having the same result.[4]
The third is high demands combined with limited decision latitude. Karasek’s demand-control model provides a framework for considering the relationship between that combination and psychological strain.[5] Being unable to change the deadline, method, or workload, while also being unable to obtain a decision, presents a problem beyond simply being busy. The model cannot by itself establish an individual’s illness or reason for leaving.
The fourth is work-related criticism accompanied by personal attacks, insults, or intimidation. Research on abusive supervision examines relationships with job dissatisfaction, exhaustion, intention to leave, and other outcomes.[6] A research construct is nevertheless distinct from a legal determination of harassment. Do not assume either that all demanding feedback qualifies or that persistent pressure inevitably produces a particular outcome. Examine the specific conduct and context.
8. Do not immediately label more frequent checking a lack of initiative
Someone who once proposed improvements, prepared materials early, automated tasks using VBA, or organized training into videos, scripts, and procedures may begin requesting detailed approval. Before explaining that change solely through motivation, look at the responses their independent action received.
They decide and hear “You acted without permission.” They ask first and hear “Think for yourself.” They coordinate and are told they made the issue bigger. They submit work and discover new conditions. When experiences like these accumulate, keeping records, asking more questions, and returning decisions to the approver can have understandable reasons.
Including necessary colleagues in an email, documenting approved content, or redirecting out-of-scope requests to the responsible department does not necessarily mean refusing work. It may be time spent avoiding later rework. Skills or information gaps may also play a part, so examine both the employee’s explanation and the actual workflow.
“Be more courageous” is not a sufficient remedy. Keep decision boundaries consistent, answer questions concretely, and accompany changes after approval with an explanation and any necessary deadline adjustment.
9. If work stalls at the supervisor, measure decision waiting time
The same employee may be able to move forward quickly after checking facts with the department responsible for a policy, yet wait a long time after consulting their direct supervisor. That difference is a useful clue when reviewing information and approval routes.
It does not, by itself, prove that the supervisor is the sole cause. Expertise, case difficulty, authority, caseload, and the need to assess risk may differ. A meaningful comparison records the request, the point at which the required information was complete, response time, and reasons for returning the work.
Separate a case’s elapsed time into the employee’s working time, time waiting for a supervisory decision, and time waiting for another department’s confirmation. This makes it harder to attribute all delay near a deadline to one person’s working style. If a decision sits unanswered for days and returns with “Finish this today,” the deadline needs to take that waiting time into account.
Changing where facts are checked does not remove formal approval or information-handling requirements. Asking the responsible department for information is different from implementing something without required authorization. Agree in advance on the usual consultation route and the alternative when no answer comes.
10. Even an outspoken employee may be unable to change the situation
Another employee may object clearly and identify specific problems, yet find that the direction remains unchanged and only the implementation question returns to them. “They cannot say what they think” is an inadequate explanation in that situation. The relevant question is how their input was considered in the decision.
Rejecting an alternative proposal is not automatically inappropriate. What matters is whether the decision-maker can explain the reason, identify who accepts the remaining risks, and provide the conditions needed to proceed. If the employee alone must explain the outcome but cannot change the direction, the burden remains.
Reasons for leaving may also overlap: family circumstances, career aims, pay and conditions, commuting, and workplace experiences can all matter. Do not dismiss a recorded “personal reason” and invent a hidden motive. Ask about work-related factors only to the extent the person wishes to discuss them. A change in facial expression or fewer smiles is not sufficient evidence of exhaustion or an intention to leave.
11. Do not confuse being easy to manage with being cooperative
Listening and cooperation matter at work. But if high ratings depend mainly on never declining vague requests, always agreeing, or managing a supervisor’s mood, practical disagreement becomes difficult. Examine whether someone is expected to provide excessive emotional caretaking without tying that expectation to gender.
When employees who have learned the job begin asking “What is the purpose?”, “Can this amount of work be completed by the deadline?”, or “Who decides?”, do not immediately interpret it as declining cooperation. Some problems become visible only as understanding of the role deepens.
Schneider’s Attraction–Selection–Attrition model is a theory for understanding organizational composition through the processes by which people are attracted to an organization, selected, and leave.[7] It does not prove that a particular supervisor retains only compliant people. Still, repeated disadvantageous treatment of employees who offer improvements is a reason to examine evaluation and retention arrangements as well as hiring. Avoid describing everyone who remains as someone who has not noticed the problems.
12. Does a discussion of work become a judgment about personality or suitability?
“This instruction conflicts with the earlier approval” or “We cannot meet this deadline with the current staffing” will not be resolved by a response consisting only of “You’re too soft,” “At your age,” or “You’re not suited to this work.”
Even when feedback on skills or performance is necessary, it should specify the behavior in question, the required standard, available support, and how improvement will be assessed. Distinguish that from sweeping judgments about someone’s character, insults in front of others, or pressure applied because they sought help.
Under Japan’s workplace power-harassment framework, three elements are considered together: conduct backed by a position of superiority, conduct beyond what is necessary and appropriate for the work, and harm to the working environment. Appropriate work instructions and guidance do not qualify, while psychological attacks and excessive demands are among the representative categories.[8] This article cannot determine whether any individual act meets that definition. Nor can Japan’s framework be used as the legal standard in other countries.
Do not interpret someone’s silence as proof that the problem has disappeared. They may fear disadvantage for speaking up, or there may be an unrelated explanation. Find out through a safe process of listening rather than assumption.
13. Seeking help outside the direct reporting line is understandable
It can be difficult to speak directly to someone who controls everyday assignments and evaluations. If that supervisor’s own conduct is the subject of the concern, there are reasons to approach HR, an internal consultation service, or a more senior manager instead.
Japan’s Ministry of Health, Labour and Welfare describes employer measures including communicating policy, providing consultation arrangements, establishing facts promptly and accurately, responding appropriately, preventing recurrence, protecting privacy, and making clear that consultation and related cooperation will not lead to disadvantageous treatment.[8] The framework does not require employees to speak only to their direct supervisor. It also does not justify promising that using a service will guarantee the desired outcome or complete anonymity.
Before seeking help, it can be useful to ask who will see the record, what will be shared, and how the information required for an investigation and any interim protection will be handled. If the direct route feels unsafe, consider another route in the organization’s rules or an appropriate external service in the relevant locality.
Keep separate records of dates and times, locations, actual words and instructions, context, your response, the effect on work, and the response you are seeking. If physical or mental effects are recorded, include only what is necessary and restrict access. Label interpretations as interpretations, and do not remove unnecessary personal information about others or confidential company material.
14. Protecting someone through a transfer and improving the original department are separate tasks
A transfer that changes contact or the evaluator may protect an employee and help them remain with the company. Their wishes, role, terms, and any potential disadvantage still need to be considered. A transfer does not guarantee recovery, and moving the person who raised a concern is not always the right response.
If a company stops at the transfer, the work definitions, workload, and management behavior may remain unchanged, leaving the next person exposed to the same burden. Protect the individual while also investigating conditions that may have contributed to the problem.
Review concentrations of resignations, transfers, and absences in the original department, along with instruction changes, reversals after approval, decision waiting time, personal remarks, and treatment after consultation. Establish facts with the parties involved and check whether the pattern continues after newcomers arrive. Moving the person who spoke up must not, by itself, become the company’s evidence of a resolution.
15. How one departure can add to the next person’s burden
A resignation leaves more than a vacancy. Unfinished cases, inquiries, training, handovers, explanations of past decisions, and coordination with other departments remain. When existing staff handle normal operations and newcomer training at the same time, they may have less time for standardization and improvement, increasing dependence on individuals.
Learning that a colleague is looking for another job may also prompt someone to reconsider their own working situation. Felps and colleagues studied relationships between coworkers’ job embeddedness and job-search behavior and an employee’s voluntary turnover.[9] “Turnover contagion” does not mean everyone will leave in the same way, nor does it place responsibility on the person who resigns. Shared workplace conditions also need consideration.
Park and Shaw’s meta-analysis found an overall negative relationship between turnover rates at group level and organizational performance.[10] The size of the association varies by context. It does not establish that every departure is bad or that turnover alone causes declining performance. The practical response is not to restrict anyone’s freedom to leave. It is to reduce the remaining employees’ workload and provide resources for handovers and support after hiring.
16. Evaluate managers by how work progresses, not just how busy they look
Lazear, Shaw, and Stanton’s “The Value of Bosses” examined differences in productivity associated with supervisors using data on technology-based service work at one company. The estimated difference in total output associated with replacing a bottom-decile boss with a top-decile boss exceeded the effect of adding one average worker to a nine-person team. Employees with better bosses were also less likely to leave.[11]
This is not a guarantee that replacing a manager will have the same effect at every company, nor is it a conclusive evaluation of any individual. The occupation, company, and measurement context matter. Nevertheless, the study gives a reason to view managers as more than communication relays and to examine whether they create conditions in which a team can work effectively.
Returning decisions, setting priorities, reducing excessive work, clarifying roles and boundaries with other departments, making it safe to raise concerns, and developing successors are behaviors to assess alongside short-term results. A manager’s visible busyness is not, on its own, a measure of management value.
17. Combine five kinds of evidence when examining a department
Movement and retention
Separate resignations, completed transfers, transfer requests, leave and long-term absence, and retention by joining cohort. When examining departments or reporting lines, distinguish reorganizations and transfers employees wanted. Report counts as well as rates, with the observation period and denominator, and avoid overinterpreting the effect of a single event in a small team.
Work handed over and knowledge lost
Look beyond the number leaving to the time required for training and handovers, concentration of core responsibilities, replacement time, and whether procedures and the reasons behind decisions remain documented. This should not become an exercise in dividing people into “valuable leavers” and everyone else.
Work definitions and changes
Record unclear ownership, unclear completion criteria, duplicate entry, verbal-only instructions, changes after approval, and work returned by other departments. Checking whether new assignments are accompanied by decisions to stop, postpone, or redistribute existing work shows whether workload adjustment actually happens.
Time to reach a decision
Examine time between requests and responses, the number of cases awaiting approval, reasons for returning work, and the owners and deadlines of unresolved decisions. Meeting length and decision counts can provide clues, but they should not encourage rushed decisions simply to improve a number. Consider case complexity and necessary checks.
Safety in raising concerns and change after intervention
Ask whether employees feel they can raise problems without disadvantage, disagree about work, use a route outside their reporting line, and avoid retaliation or disadvantage after seeking help. More consultations can reflect greater trust in the service as well as more problems. Zero reports should not automatically be treated as a good result.
These records are not a surveillance or blame-finding checklist. Limit access, gather only necessary information, and avoid identifying individuals from small-group responses. Stricter numerical targets for managers alone may create incentives to suppress consultations or transfers. Combine numbers with interviews and work records.
18. What to fix first: workload, conditions, decisions, and consultation
First, consider protection from immediate excessive demands or intimidation, and review the work already in progress. When adding work, decide at the same time what will stop, be postponed, be reassigned, change in scope, or receive a renegotiated deadline. If nothing can be reduced, a more senior decision-maker needs to take responsibility for that choice and the resources it requires.
Next, put the assignment’s conditions in a short written brief. Include its purpose, deliverable, mandatory and optional requirements, deadline, priority, responsibility boundaries, the employee’s decision authority, review points, final decision-maker, and responsibility for exceptions. Not everything needs a detailed manual. What matters is avoiding areas for which nobody will make a decision.
Set response expectations for managers too. Agree on appropriate timeframes for ordinary and urgent cases, expressed in business days or hours as the work requires. Specify how an inability to decide will be communicated, who can make an alternative decision, and how waiting will affect deadlines. Lack of a response must not silently become permission to proceed without authorization.
Evaluate consistency of instructions, workload reduction, onboarding support, handovers, responses to employees’ views, and recurrence prevention alongside short-term performance. Protect personal choice and safety so that attempts to improve retention do not obstruct transfers or resignations.
Consultation staff are not there to judge someone’s age, character, or toughness. They should establish the chronology, facts, effect on work, wishes of the person seeking help, and need for interim protection. Interview involved parties separately and explain information-sharing boundaries and measures against retaliation. Do not decide in advance that the supervisor is right or the employee mistaken without the necessary inquiry.
Finally, check whether improvement lasts. Reviews after three and six months, for example, can revisit instruction changes, decision waiting time, workload, and treatment after consultation. Those are practical examples of timing, not a universally effective research-prescribed schedule. Changes in everyday behavior matter more than the number of events or training sessions held.
19. What individuals can do: clarify conditions and widen safe options
A structural problem cannot be solved through one person’s communication style or endurance alone. Where it is safe, turn vague requests into concrete questions. For example: “Before I start, could we confirm the scope, reference date, submission date, mandatory items, and final decision-maker?”
Keep approved content in a form that can be checked later. A useful formulation is: “I’ll use the previously approved content as the baseline. If it changes, I’d like to confirm what changed, why, and the impact on the deadline.” A record is not an absolute defense, but it can help clarify differences in understanding.
For additional work, identify the choice that needs to be made: “Adding this to A and B will affect the deadline for C. Which should take priority, and which should be postponed?” You do not need to make an unconditional promise to handle everything before examining the burden.
For sales calls, deliveries intended for another department, visitors without a clear contact, facilities inquiries, or unresolved meeting issues, identify the entry point and responsible department. Do not neglect an urgent safety issue; separate the immediate response from permanent ownership. The aim is to prevent work accepted as a favor from becoming a standing assignment without a formally responsible owner.
In records, distinguish actual words and instructions; your checks, alternatives, and submissions; effects such as delay and rework; interpretations of the other person’s intentions; and your own feelings. Do not present an inference as a fact, and store necessary material securely in accordance with applicable rules. Secret recording or taking documents outside the company is not being recommended as a routine response.
Rather than trying to win an argument on the spot, it may be better to confirm the issue in writing, share it only with necessary people, use another consultation route, or consider a transfer or job change. If discussion itself raises fears of harm or retaliation, prioritize advice from a trusted channel over direct confrontation. What is safe and appropriate depends on the organization, local systems, and the person’s circumstances.
20. Before increasing recruitment, check whether the same cycle is repeating
One pattern to investigate begins when experienced employees fill gaps in undefined work, the company mistakes their extra effort for a functioning process, and assignments keep being added. Conditions do not change when problems are raised. Increased checking becomes a criticism of the employee’s attitude, and a new recruit fills the gap after a transfer or resignation. If the conditions remain, the next person may face the same burden.
This is a hypothesis to test, not a predetermined finding. Examine pay, working hours, business changes, commuting, personal preferences, and managers’ own workload or lack of authority as well. Reports of the same problem from several people do not remove the need to establish individual facts and consider evidence against an explanation.
Recruitment may be necessary, but replacement alone should not be treated as the solution. Respect the explanations of people who leave, make it safe for those who remain to speak, and change the conditions needed to make the work workable. That is the starting point for addressing the risk of another departure.
Conclusion: Ask whether people can stay after learning the job
Even with company-wide turnover of 4%, a particular department may be losing accumulated experience. Conversely, extensive replacement alone does not establish a problem with a manager or workplace. Align time periods and denominators, then examine staff movement, work definitions, decision waiting time, workload, and responses to consultation together.
When employees start checking more and avoiding decisions, do not stop at “They lack initiative.” Are they rejected for acting independently, criticized for asking, left without approval, and still expected to bear responsibility for results? Investigating those conditions offers a more useful route to improvement than speculating about personality.
Bringing people in and enabling them to keep working after they have learned the job are different achievements. The goal is not to make the numbers look better. It is to ensure that decisions are made by those responsible, unnecessary work is reduced, people who seek help are protected, and the same problems do not keep recurring.
References (11)
[1] Bauer, T. N., Bodner, T., Erdogan, B., Truxillo, D. M., & Tucker, J. S. (2007). Newcomer adjustment during organizational socialization: A meta-analytic review of antecedents, outcomes, and methods. Journal of Applied Psychology, 92(3), 707–721. https://pubmed.ncbi.nlm.nih.gov/17484552/
[2] Japan Institute for Labour Policy and Training (March 31, 2022). Research Report No. 215, “Job Changes, Skills Development and Career Formation among the Middle-Aged” (『ミドルエイジ層の転職と能力開発・キャリア形成』); and Makoto Fujimoto (July 2022), Labour Policy Forum presentation, “Job Changes, Use of Skills and Career Formation: Considering Job Changes among the Middle-Aged” (『転職と能力発揮・キャリア形成~ミドルエイジ層の転職から考える~』). English titles are descriptive translations. https://www.jil.go.jp/institute/reports/2022/documents/0215.pdf / https://www.jil.go.jp/event/ro_forum/20220726/resume/01-kenkyu-fujimoto.pdf
[3] Jackson, S. E., & Schuler, R. S. (1985). A meta-analysis and conceptual critique of research on role ambiguity and role conflict in work settings. Organizational Behavior and Human Decision Processes, 36(1), 16–78. https://doi.org/10.1016/0749-5978(85)90020-2
[4] Tubre, T. C., & Collins, J. M. (2000). Jackson and Schuler (1985) revisited: A meta-analysis of the relationships between role ambiguity, role conflict, and job performance. Journal of Management, 26(1), 155–169. https://doi.org/10.1177/014920630002600104
[5] Karasek, R. A. (1979). Job Demands, Job Decision Latitude, and Mental Strain: Implications for Job Redesign. Administrative Science Quarterly, 24(2), 285–308. https://doi.org/10.2307/2392498
[6] Bhattacharjee, A., & Sarkar, A. (2024; online first 2022). Abusive supervision: a systematic literature review. Management Review Quarterly, 74, 1–34. https://pmc.ncbi.nlm.nih.gov/articles/PMC9376912/
[7] Schneider, B. (1987). The People Make the Place. Personnel Psychology, 40(3), 437–453. https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1744-6570.1987.tb00609.x
[8] Japan’s Ministry of Health, Labour and Welfare. Employer guidance on harassment prevention (『事業主が講ずべきハラスメント対策パンフレット』) and the “Akarui Shokuba Oendan” workplace support portal. Workplace power harassment in Japan and employers’ prevention measures. https://www.mhlw.go.jp/content/11900000/001338359.pdf / https://www.no-harassment.mhlw.go.jp/foundation/harassment_list/power-hara/
[9] Felps, W., Mitchell, T. R., Hekman, D. R., Lee, T. W., Holtom, B. C., & Harman, W. S. (2009). Turnover Contagion: How Coworkers' Job Embeddedness and Job Search Behaviors Influence Quitting. Academy of Management Journal, 52(3), 545–561. https://doi.org/10.5465/AMJ.2009.41331075
[10] Park, T.-Y., & Shaw, J. D. (2013). Turnover rates and organizational performance: A meta-analysis. Journal of Applied Psychology, 98(2), 268–309. https://pubmed.ncbi.nlm.nih.gov/23244224/
[11] Lazear, E. P., Shaw, K. L., & Stanton, C. T. (2015). The Value of Bosses. Journal of Labor Economics, 33(4), 823–861. NBER working paper (2012), published version (2015). https://www.nber.org/papers/w18317



