Ethical dilemma analysis tool for HR professionals and Level 7 students
Ethical dilemma analysis tool for HR professionals and Level 7 students
When there is no easy “right” answer
An organisation discovers that an employee has underperformed for several months. The employee is also experiencing significant personal difficulties. The line manager wants immediate dismissal because performance is affecting the team. The HR adviser believes the organisation should offer additional support and a reasonable opportunity to improve. Senior leaders are concerned about productivity and consistency. Other employees believe the situation is unfair because colleagues who perform poorly appear to receive different treatment.
What should HR do?
There may be no answer that satisfies every legitimate interest. The decision involves fairness, compassion, consistency, performance, organisational sustainability, employee rights, managerial authority, evidence and professional responsibility. A compassionate response may be criticised as inconsistent. An immediate dismissal may protect short-term output but fail to respect procedural fairness or relevant personal circumstances. A long process may support the employee but impose costs on colleagues and customers.
This is the kind of situation for which an ethical dilemma analysis tool is useful. An ethical dilemma is not simply a situation in which something feels wrong. It is a situation involving competing values, duties, interests, rights or legitimate stakeholder claims where the appropriate course of action is not immediately obvious. The aim is not to produce a mathematically perfect answer. It is to make a decision that is reasoned, evidence-informed, proportionate, transparent, professionally responsible and more ethically defensible than the available alternatives.
The tool in this resource moves the analyst through the following sequence:
Situation → ethical issue → stakeholders → values → options → ethical frameworks → evidence → consequences → duties and rights → professional standards → decision → justification → review.
It is designed for assignments, case studies, reflective work, professional practice, employee relations, people management and strategic decision-making. It also provides a common language for constructive challenge: HR can disagree with a manager without reducing the disagreement to personal preference.
What is an ethical dilemma?
Core definitions
| Concept | Working definition for HR analysis |
|---|---|
| Ethics | Reasoned standards and practices concerning how people ought to act and how institutions ought to treat those affected by their decisions. Ethics concerns conduct, relationships, responsibilities, consequences, rights and character. |
| Morality | A person’s or community’s beliefs about right and wrong. Morality may inform ethical reasoning, but a private moral reaction is not by itself a complete professional analysis. |
| Values | Principles or commitments regarded as important, such as dignity, fairness, autonomy, compassion, honesty, inclusion, efficiency or accountability. |
| Ethical issue | A morally significant feature of a decision, such as possible harm, unequal treatment, deception, coercion, privacy intrusion or a failure of responsibility. |
| Ethical dilemma | A situation in which two or more legitimate values, duties, rights, interests or stakeholder claims pull in different directions, making the best course of action contestable. |
| Ethical decision | A decision reached through structured consideration of facts, uncertainty, stakeholders, values, rights, duties, consequences, evidence and professional responsibilities. |
| Professional responsibility | The responsibilities arising from a professional role, including competence, integrity, confidentiality, courage, accountability, respect and the duty to challenge harmful or indefensible practice. |
Ethical problem, ethical dilemma and legal problem
An ethical problem exists where a relatively clear ethical obligation is being ignored—for example, a manager knowingly falsifying an absence record. An ethical dilemma exists where legitimate obligations conflict—for example, whether to disclose sensitive information to protect a group when confidentiality also matters. A legal problem concerns a legal requirement, prohibition, entitlement or remedy. A dilemma may involve all three.
Legal does not automatically mean ethical, and ethical does not mean ignoring legal requirements.
Law normally establishes a minimum enforceable standard, although some legal rules are also powerful expressions of ethical values. Ethics asks additional questions: Is the process respectful? Is the action proportionate? Has power been used responsibly? Were less harmful alternatives available? Can the decision be justified to those affected? The Markkula Center for Applied Ethics similarly cautions that ethics is not the same as feelings, law, accepted norms or science; science provides evidence, but evidence alone does not decide what ought to be done.[7]
Why ethical dilemmas are difficult in hr
HR decisions are ethically demanding because they affect people’s income, identity, security, dignity, opportunity, health and ability to participate in working life. Employment is also a relationship of dependency: an organisation often controls access to work, information, progression, pay and formal voice. The resulting power imbalance means that apparent consent may be constrained and that a neutral-looking process can have unequal effects.
HR professionals may owe responsibilities to employees, managers, leaders, the organisation, colleagues, customers, owners, regulators, professional bodies and wider society. These responsibilities may be mutually supportive, but they may also conflict. For example, a decision that protects business continuity may impose severe burdens on a vulnerable workforce; a decision that protects an individual may create perceived inconsistency; and a policy applied identically to everyone may disadvantage people whose circumstances are materially different.
| Source of difficulty | Ethical question |
|---|---|
| Power | Who defines the problem, controls the evidence and can challenge the outcome? |
| Dependency | Can the employee genuinely refuse, consent or speak freely? |
| Confidentiality | What information must be protected, and what explanation is owed to others? |
| Uncertainty | What is known, what is inferred and what remains untested? |
| Competing interests | How should employee, team, customer and organisational interests be weighed? |
| Time pressure | Does urgency justify reduced consultation or does it make safeguards more important? |
| Professional role | Is HR a neutral administrator, employee advocate, organisational adviser, guardian of process, or all of these in different contexts? |
| Systems effects | How might the decision alter manager behaviour, culture, trust and future cases? |
The CIPD describes ethical practice as building trust by role-modelling ethical behaviour and applying principles and values consistently in decision-making.[1] That expectation is demanding precisely because ethical practice includes both the decision and the way it is implemented. Even when not everyone can receive the preferred outcome, people can be treated with dignity, given a fair process and told enough about the reasoning to understand what happened.
The ethical dilemma analysis tool
The tool has thirteen linked stages. Do not treat it as a tick-box exercise. Each stage should sharpen the next one.
| Stage | Key question | Output |
|---|---|---|
| 1. Situation | What is happening? | A neutral description |
| 2. Ethical issue | Why is this morally significant? | The value conflict or risk |
| 3. Decision | What must be decided now? | A precise decision question |
| 4. Facts and uncertainty | What is established and what is assumed? | Fact/assumption log |
| 5. Stakeholders | Who is affected, heard and vulnerable? | Stakeholder map |
| 6. Values | Which commitments support each option? | Competing-values table |
| 7. Options | What could be done, including mitigations? | At least three realistic options |
| 8. Frameworks | What do consequentialist, duty, virtue and justice lenses reveal? | Comparative analysis |
| 9. Evidence | What evidence would confirm, challenge or refine the analysis? | Evidence plan and quality assessment |
| 10. Duties and standards | What rights, duties, law, policy and professional standards constrain the decision? | Compliance and responsibility map |
| 11. Consequences and proportionality | What harms, benefits, reversibility and safeguards exist? | Options matrix and proportionality test |
| 12. Judgement and implementation | Which option is most defensible and how will it be enacted? | Reasoned recommendation and safeguards |
| 13. Review | What will be monitored, learned and reconsidered? | Review measures and trigger points |
The sequence is iterative. New evidence may change the stakeholder map; stakeholder voice may reveal that the original decision question was too narrow; and implementation risks may require a new option rather than a simple choice between “yes” and “no.”
Defining the decision
Start by replacing the vague question “Is this situation ethical?” with a decision question that identifies the action, the purpose and the tension.
Should the organisation proceed with X, despite Y, in order to achieve Z, and subject to which safeguards?
For the opening case, a weak question is “Should we dismiss the employee?” A stronger question is: “Should the organisation begin a formal capability process now, or provide a defined period of supported improvement first, given the performance evidence, the employee’s circumstances, the team impact and the need for consistent treatment?”
A properly defined decision has five features. It identifies the decision-maker; states what is genuinely within scope; separates the decision from the outcome hoped for; recognises the competing concern; and leaves room for alternatives. A good decision statement prevents analysts from smuggling a preferred answer into the question.
Decision-definition prompts
| Prompt | Example |
|---|---|
| Who decides? | Business leader with HR advice and, where relevant, employee representation. |
| What is the immediate decision? | Whether to implement a supported improvement process, formal action, or another route. |
| What objective is sought? | Reliable performance and a fair, sustainable process. |
| What constraint exists? | Team workload, confidentiality, policy, evidence quality and applicable law. |
| What must not be assumed? | That support will work, that dismissal is inevitable, or that consistency means identical treatment. |
Facts vs assumptions
Ethical analysis becomes weaker when assumptions are presented as facts. The analyst should create a live evidence log and label each statement as established, reported, inferred, contested or unknown.
| Established facts | Assumptions and questions |
|---|---|
| What has been documented? | Is the documentation complete and reliable? |
| What has been directly observed? | Could the behaviour have another explanation? |
| What policy or contractual process applies? | Has the policy been applied consistently? |
| What impact has been measured? | Is the claimed impact causal or merely correlated? |
| What has the employee said? | Has the employee had a meaningful opportunity to respond? |
| What support has been offered? | Was it accessible, appropriate and understood? |
A useful evidence register records the source, date, relevance, reliability, possible bias and what would change the conclusion. A manager’s statement that “the team is suffering” may be important stakeholder evidence, but it is not the same as absence data, quality data, customer evidence or an independently reviewed workload assessment.
Fact-checking questions
Ask: What would a sceptical reader say is missing? Which facts benefit the preferred option, and which challenge it? Are different standards of proof being used for different stakeholders? Does the analysis disclose material uncertainty instead of concealing it? Is confidential information being used only for a legitimate purpose and shared only with those who need it?
Stakeholder analysis
Stakeholder analysis is more than a list of interested parties. It asks whose interests are affected, whose voice is missing, who has power to influence the process and who bears the consequences.
| Stakeholder | Wants | Needs or rights | Risks faced | Power and voice | Ethical question |
|---|---|---|---|---|---|
| Directly affected employee | Dignity, fair treatment, clarity, opportunity | Respect, relevant explanation, meaningful process | Loss of income, stigma, wellbeing harm | Often lower formal power | Has the person been heard and treated as an end, not merely a means? |
| Other employees | Workable standards and fair workload | Consistency, safety, trust | Resentment, overload, fear | Collective voice may vary | Are their interests being used fairly rather than rhetorically? |
| Line manager | Performance and authority | Support, clarity, accountability | Team failure, criticism | Direct control | Is managerial convenience being mistaken for organisational necessity? |
| HR | Ethical, lawful and effective process | Independence, competence, support | Complicity, loss of trust | Process influence | Has HR challenged rather than merely processed? |
| Senior leaders | Sustainability and reputation | Accurate advice and risk visibility | Financial and reputational risk | High | Are short-term results crowding out long-term legitimacy? |
| Customers and service users | Reliable service | Quality and continuity | Harm from service failure | Indirect | Are customer effects evidenced? |
| Representatives or unions | Voice and protection | Consultation and information | Loss of influence | Variable | Is consultation genuine or tokenistic? |
| Regulators and society | Compliance and responsible conduct | Public interest | Systemic harm | External | Does the decision reinforce or undermine trustworthy institutions? |
Use a power–interest map, but do not equate low power with low importance. A person with little organisational authority may experience the greatest harm. Ask: Who gets heard? Who can challenge the decision? Who is vulnerable? Who benefits? Who bears the cost?
Employee involvement does not automatically make a decision ethical. Consultation can be tokenistic, framed by management, inaccessible or disconnected from influence. The ethical quality of voice depends on purpose, representation, information, independence, feedback and the extent to which participation can alter the outcome.
Competing values
The point is not to list values. The point is to show where values conflict and what each value requires in the specific context.
| Value | What it requires here | Who benefits? | What conflict exists? |
|---|---|---|---|
| Fairness | A process and outcome that can be justified by relevant reasons | All parties | Consistency may conflict with responsiveness to individual circumstances |
| Dignity | Respectful treatment and no humiliating use of personal information | Directly affected person and workforce | Performance pressure may encourage instrumental treatment |
| Compassion | Serious attention to vulnerability and preventable harm | Vulnerable employees and culture | Compassion may be portrayed as leniency or inconsistency |
| Accountability | Clear standards and consequences for failure | Team, customers and organisation | Accountability may conflict with support or rehabilitation |
| Autonomy | Honest information and meaningful participation | Employees and representatives | Managerial authority limits freedom to choose |
| Confidentiality | Restrict sensitive information to legitimate need | Individuals and trust | Others may demand transparency about unequal treatment |
| Inclusion | Consideration of relevant differences and barriers | Under-represented or disadvantaged groups | Uniform rules may produce unequal effects |
| Efficiency | Avoid waste and deliver reliable outcomes | Customers and organisation | Efficiency can externalise costs to employees |
| Sustainability | Protect future capability, trust and viability | Current and future stakeholders | Immediate cost reduction may undermine resilience |
| Honesty | No misleading explanation or selective disclosure | Everyone affected | Full disclosure may conflict with privacy or investigation integrity |
A strong analysis explicitly states the tension: “The case is not a simple choice between fairness and compassion. It requires the organisation to decide what fairness means when relevant personal circumstances affect performance, and whether support is a fair accommodation rather than an unfair exception.”
The four ethical frameworks
Ethical frameworks are lenses, not answer machines. Each illuminates some features and can obscure others.
1. Utilitarianism and consequentialism
The central question is: Which option produces the greatest overall benefit and least harm for those affected? Analyse the likely benefits, harms, probability, severity, timing, distribution and reversibility of consequences. Include short-term, long-term and unintended consequences.
| Strengths | Limitations | HR example |
|---|---|---|
| Makes stakeholder impact visible; supports comparison; recognises indirect and future effects. | Benefits and harms may be difficult to measure; minorities can be sacrificed for aggregate gain; forecasts are uncertain. | Monitoring might increase output, but privacy loss, stress, gaming and turnover may outweigh measured productivity gains. |
A consequentialist analysis of the performance case may consider team workload, customer effects, the employee’s chance of improvement, turnover, precedent and trust. It must not assume that a financial benefit is automatically the greatest good.
2. Deontology and duty ethics
The central question is: What duties, rights and principles should guide the decision regardless of the outcome? Consider respect for persons, truthfulness, promises, consistency, non-discrimination, confidentiality, due process and professional obligations.
| Strengths | Limitations | HR example |
|---|---|---|
| Protects dignity and rights; resists treating people merely as means; gives clear constraints. | Duties may conflict; rigid rules can neglect consequences and context; rights may be contested. | An employee should not be subjected to covert monitoring merely because it is efficient if duties of transparency, privacy and respect are not satisfied. |
In performance management, a duty-based approach asks whether the employee was told the standard, given a genuine opportunity to respond and assessed using reasons that could be applied to comparable cases.
3. Virtue ethics
The central question is: What would a good and practically wise professional do, and what kind of organisation are we becoming through this decision? Relevant virtues include integrity, courage, honesty, fairness, compassion, humility, prudence and practical wisdom.
| Strengths | Limitations | HR example |
|---|---|---|
| Develops character, culture and professional courage; handles ambiguity and relationships well. | Virtues can be interpreted differently; good intentions do not guarantee good outcomes; structures may overpower character. | A courageous HR adviser respectfully challenges an immediate dismissal while also acknowledging the team’s legitimate needs. |
Virtue ethics is not a call to be “nice.” Compassion without accountability may be irresponsible; courage without humility may become self-righteousness; loyalty without integrity may become complicity.
4. Justice and fairness
The central question is: Is the decision fair, equitable and justifiable? Consider distributive justice (outcomes), procedural justice (process), interactional justice (respectful treatment) and informational justice (adequate explanation). Distinguish equality—treating people the same—from equity—taking relevant differences into account to achieve fair treatment.
| Strengths | Limitations | HR example |
|---|---|---|
| Directly addresses consistency, voice, bias, distribution and process legitimacy. | People may disagree about the relevant standard; equal outcomes are not always fair; justice may conflict with efficiency or care. | A redundancy process is not fair merely because the same formula is applied to everyone if the criteria have disproportionate, avoidable effects on one group. |
Justice asks both “Was the result fair?” and “Was the route to the result fair?” A fair outcome produced through humiliation or hidden criteria remains ethically defective.
When ethical frameworks disagree
Consider employee monitoring. A utilitarian may see productivity gains and ask whether safeguards reduce privacy harms. A deontologist asks whether consent, transparency and respect for privacy have been honoured. Virtue ethics asks what a trustworthy employer would do and whether surveillance expresses prudence or distrust. Justice asks whether monitoring is proportionate, consistently governed and more burdensome for particular groups or roles.
The frameworks may recommend different options. That is not a defect. Ethical analysis is not about selecting the framework that gives the answer already wanted. It is about exposing the dimensions of the dilemma, testing assumptions and explaining why one combination of reasons is more defensible in context.
A useful comparison table is:
| Lens | Main question | Likely insight | Blind spot to test |
|---|---|---|---|
| Consequences | What produces the best balance of benefit and harm? | Aggregate impact and unintended effects | Who may be sacrificed or made invisible? |
| Duties and rights | What must never be done, and what is owed? | Dignity, privacy, voice and constraints | What happens when duties conflict? |
| Virtue | What would a good professional or organisation do? | Character, courage and practical wisdom | Are good intentions supported by evidence and systems? |
| Justice | Who receives benefits and burdens, and was process fair? | Equity, consistency, voice and legitimacy | Which standard of fairness is being used? |
The ethical dilemma matrix
Use this matrix before writing a recommendation.
| Question | Analysis |
|---|---|
| What is the decision? | |
| What are the established facts? | |
| What is uncertain or contested? | |
| Who are the stakeholders? | |
| Who is vulnerable or under-heard? | |
| What values conflict? | |
| What rights and duties apply? | |
| What are the realistic options? | |
| What evidence is available and how strong is it? | |
| What would each framework suggest? | |
| What professional standards apply? | |
| What legal constraints exist? | |
| What are the benefits and harms? | |
| What risks and unintended consequences exist? | |
| What is the fairest option? | |
| What is the most proportionate option? | |
| What is the most defensible decision? | |
| How should it be implemented? | |
| What safeguards and voice mechanisms are required? | |
| How will the decision be reviewed? |
Legality vs ethics
“Is It Legal?” Is Only the First Question
A useful sequence is:
Legal minimum → organisational policy → professional standards → ethical responsibility → best defensible practice.
The legal minimum is essential but may not settle how a process should be designed or implemented. Organisational policy may exceed law, while professional standards may require courage, competence and integrity beyond minimum compliance. Ethical responsibility asks what is owed to people and society; best defensible practice combines legal compliance, sound evidence, fair process, proportionality and accountability.
| HR topic | Legal/regulated question to verify | Wider ethical questions |
|---|---|---|
| Monitoring | Is there a lawful and transparent basis for processing data? | Is monitoring necessary, proportionate, trusted and the least harmful means? |
| Redundancy | Are consultation, selection and notice requirements met? | Are burdens fairly distributed and alternatives genuinely considered? |
| Performance | Are process and reasons consistent with applicable law and policy? | Has context, support, dignity and voice been considered? |
| Recruitment | Are equality and data-protection obligations respected? | Does the method reproduce bias or conceal arbitrary judgement? |
| Pay | Are contractual, equality and reporting duties met where applicable? | Is the reward system justifiable, transparent and sustainable? |
| Absence | Is sensitive information handled lawfully? | Is support genuine, and are health or disability-related differences treated fairly? |
| AI | Are data, equality, transparency and automated-decision rules satisfied? | Who is accountable when the system is wrong, biased or inscrutable? |
| Whistleblowing | Is the report handled under applicable protection and investigation rules? | Can the person speak without retaliation, and is the public interest protected? |
For UK readers, the Equality Act 2010 legally protects people from discrimination in the workplace and wider society.[6] The ICO’s worker-monitoring guidance concerns monitoring and data protection, but warns that it is under review following the Data (Use and Access) Act; it should therefore be checked for currency before action.[5] The resource does not provide jurisdiction-specific advice. Readers should consult current legislation, regulators, ACAS, specialist advisers and organisational procedures.
The cipd profession map and ethical practice
The current CIPD Profession Map identifies ethical practice as a core behaviour and describes it as building trust by role-modelling ethical behaviour and applying principles and values consistently in decision-making.[1] Its wider behavioural expectations include professional courage and influence, valuing people, working inclusively and situational decision-making.[1] The CIPD Code of Conduct and Ethics sets expectations for members and is organised around five virtues: positive and active impact on working lives; civic virtue and stewardship; good character; professional service and competence; and personal responsibility.[2]
These standards are highly relevant to ethical dilemma analysis. Principles-led practice asks whether the decision is anchored in defensible values. Evidence-based practice asks whether claims and proposed interventions are supported by reliable evidence. Professional courage and influence asks whether HR has challenged an unsafe, unfair or misleading direction. Valuing people asks whether people are treated with dignity. Working inclusively asks whether differences, barriers and unequal effects have been considered. Situational decision-making asks whether the response is sensitive to context rather than mechanically uniform.
The Profession Map provides professional expectations; ethical frameworks provide different ways of reasoning through the dilemma.
The Profession Map does not itself resolve every dilemma. A professional standard can tell us to act ethically, competently and responsibly without deciding whether a particular monitoring system is proportionate or whether a supported performance process is fair. That is why the Map should be used alongside the four frameworks, evidence, stakeholder voice and critical judgement.
For broader study, connect this resource to The CIPD Profession Map Explained for Level 7 Assignments and Is HR a Profession? The Debate Level 7 Markers Want You to Engage With. The publisher should insert verified internal URLs rather than relying on invented links.
Evidence-based ethical decision-making
Ethical reasoning and evidence-based practice should work together. Intuition can identify that something is troubling, but intuition alone may reflect bias, status, fear or personal preference. The CIPD’s current evidence-based practice guide identifies four sources: professional expertise, scientific literature, organisational data and stakeholder views.[3] It recommends six steps: ask critical questions; acquire evidence; appraise it; aggregate it; apply it; and assess the outcome.[3]
| Evidence source | What it can contribute | Critical question |
|---|---|---|
| Scientific literature | General knowledge about likely effects and mechanisms | Is the research relevant, current and methodologically credible? |
| Organisational data | Local patterns, outcomes, costs and disparities | What is measured, omitted or distorted by the data-generating process? |
| Stakeholder views | Experience, meaning, concerns and practical knowledge | Who was heard, who was absent and how might power shape responses? |
| Professional expertise | Contextual judgement and implementation knowledge | Is the expert transparent about uncertainty and conflicts of interest? |
Ask not only “What evidence supports my preferred option?” but also “What evidence challenges it?” A manager’s confidence is not proof. A dashboard’s precision is not validity. A statistically different outcome is not automatically discriminatory, and a lack of measured difference is not proof of fairness.
Evidence-quality prompts
Record provenance, relevance, recency, completeness, reliability, bias, uncertainty, privacy implications and the consequences of acting on a false positive or false negative. Where evidence is weak but a decision cannot wait, state the uncertainty, choose reversible safeguards where possible and set a review trigger.
Connect this section to Evidence-Based Practice in HRM: What It Means and Why It Matters. Evidence should inform ethical judgement, not displace it: a finding that an intervention increases productivity does not decide whether its burdens are acceptable.
Professional judgement
Evidence and frameworks do not automatically produce a single answer. HR professionals must exercise judgement under uncertainty, competing interests, incomplete information, time pressure, organisational power and accountability. Professional judgement is not the same as personal preference.
A defensible professional judgement should be reasoned, evidence-informed, ethically considered, context-sensitive, transparent, proportionate and accountable. It should explain what was decided, why the alternatives were rejected, what uncertainty remains, who approved the decision, what safeguards apply and how the outcome will be reviewed.
A useful judgement statement is:
“On the current evidence, option B is the most ethically defensible because it protects the strongest duties, reduces foreseeable harm, gives affected stakeholders meaningful voice and remains proportionate to the objective. It is not risk-free. Its safeguards are X and Y; if review measure Z indicates harm or unequal impact, the decision will be reconsidered.”
Professional courage means challenging a preferred management answer without exaggerating certainty or pretending HR has no organisational responsibilities. It also means escalating when the decision exceeds one’s competence, independence or authority.
Power and ethical decision-making
Power determines who defines the problem, what counts as evidence, who is heard, who can challenge the decision, who bears the consequences and who benefits. It operates through hierarchy, expertise, access to information, job security, social status, professional language, technology and control of process.
Power is ethically relevant in investigations, discipline, redundancy, whistleblowing, performance ratings, recruitment, promotion and pay. An apparently neutral process can produce unequal outcomes when the baseline reflects historical advantage, when voice is constrained or when the decision-maker controls both the criteria and the interpretation of results.
Use these questions:
| Power question | Application |
|---|---|
| Who framed the issue? | Has the problem been defined as an individual failure when it may be systemic? |
| Who is absent? | Are contractors, disabled employees, remote workers, representatives or future employees missing? |
| Who controls evidence? | Can the affected person inspect, correct or contextualise the record? |
| Who can dissent safely? | Are there independent routes for challenge and escalation? |
| Who bears risk? | Are reputational, financial, emotional and career costs distributed fairly? |
| What becomes normal? | Will the decision make surveillance, silence, favouritism or overwork more acceptable? |
Power analysis should not assume that managers are always wrong or employees always right. It should make asymmetry visible and require stronger safeguards where the affected person has limited capacity to challenge the organisation.
Fairness is not always equality
Equality means treating people the same according to a stated standard. Equity means taking relevant differences into account so that treatment and outcomes can be fair. Equal treatment can be unfair when people face materially different barriers; different treatment can be unfair when it is based on irrelevant preference or bias.
| Form of fairness | Question | HR example |
|---|---|---|
| Distributive | Are benefits and burdens fairly allocated? | Are redundancy burdens concentrated in one group without defensible reason? |
| Procedural | Was the process consistent, impartial, evidence-based and open to voice? | Were comparable employees given comparable opportunities to respond? |
| Interactional | Were people treated with respect and dignity? | Was difficult feedback delivered privately and without humiliation? |
| Informational | Was an honest, adequate explanation provided? | Can an employee understand the basis of a decision without receiving another person’s confidential data? |
| Equality | Were relevant cases treated alike? | Are criteria applied consistently? |
| Equity | Were relevant differences considered? | Was an accessible process provided where a disability or circumstance made the standard route inappropriate? |
The ethical question is not “Did everyone receive exactly the same thing?” It is “Was the difference in treatment based on relevant reasons, applied consistently and justified to those affected?”
Confidentiality and transparency
HR needs confidentiality to protect privacy, investigation integrity and trust. Employees also need transparency to understand decisions, challenge errors and assess whether the organisation is fair. The ethical task is not maximum disclosure or maximum secrecy; it is appropriate transparency.
Ask: What information is necessary to explain the decision? Whose confidentiality is protected? Who needs to know? Can the organisation describe the process, criteria, safeguards and review without revealing personal details? Is confidentiality being used to protect people or to avoid accountability?
In disciplinary and grievance matters, confidentiality should not become a reason to deny a person a meaningful opportunity to respond. In whistleblowing, privacy should coexist with protection from retaliation and credible investigation. In pay and restructuring, aggregate explanations may increase fairness without exposing individual data. In AI decisions, transparency should include the purpose, data categories, human oversight, routes to challenge and limits of the system.
Employee voice
Employee voice includes consultation, participation, representation, trade-union engagement, surveys, collective forums, grievance routes and direct challenge. Voice is ethically valuable because affected people possess experience and because participation can correct information asymmetry.
However, involvement is not automatically ethical. Voice can be tokenistic, selective, unsafe, inaccessible or disconnected from influence. A responsible process explains the decision space, gives enough information to contribute, includes under-heard groups, records dissent, responds to what was said and explains what changed or why it did not.
Ask not only “Were employees consulted?” but “Could their participation make a difference, and could they disagree without unreasonable risk?”
Proportionality
Use proportionality when an action imposes a burden in pursuit of a legitimate objective.
- Necessity: Is the action genuinely necessary, or is the problem being overstated?
- Effectiveness: Is it reasonably likely to achieve the intended objective?
- Least harmful alternative: Is there a less intrusive or less harmful way to achieve the same objective?
- Fairness: Are burdens distributed fairly, and are relevant differences considered?
- Transparency: Can the reasoning be explained without misleading or exposing protected information?
- Accountability: Who owns the decision and the consequences?
- Review: Can the decision be reconsidered, corrected or stopped?
Application: employee monitoring
Before introducing monitoring, define the problem and test whether workload, unclear goals, inadequate resources or poor management explain the performance issue. Compare aggregated, limited and role-specific data with continuous individual surveillance. Set retention, access, notification, challenge and deletion safeguards. Review whether monitoring changes behaviour through gaming, stress or presenteeism.
Application: AI recruitment
Test whether the tool adds valid value beyond structured human assessment. Examine data provenance, validation, error rates, disparate impacts, accessibility, explainability and human override. Use the least intrusive data and ensure candidates can obtain meaningful information and challenge material errors where applicable.
Ethical dilemma case studies
Each case below uses the full tool. The recommendations are examples of defensible reasoning, not universal answers.
Case 1 — Performance management and personal difficulties
Facts and uncertainties. Performance has fallen for several months and team workload has increased. The employee reports significant personal difficulties. It is uncertain whether expectations were clear, whether support was offered, whether comparable cases were handled consistently and whether the performance gap is caused by individual capability, workload, management, health, resources or role design.
Stakeholders and values. The employee, team, manager, HR, leaders and customers have legitimate interests. Dignity, compassion, fairness, accountability, consistency, wellbeing and organisational performance conflict.
Frameworks and evidence. Consequentialism compares team impact, recovery prospects, turnover and trust. Duty ethics requires clear standards, honest process, respect and confidentiality. Virtue ethics asks whether HR shows courage and practical wisdom. Justice asks whether the process is fair and whether relevant circumstances have been considered. Evidence should include objectives, quality data, workload, support records, the employee’s account and comparable cases.
Legal and professional considerations. Check current capability, equality, absence, disability, privacy and contractual requirements for the jurisdiction. The CIPD Profession Map supports ethical, evidence-based and courageous practice.[1] Avoid diagnosing or disclosing private information beyond legitimate need.
Options and consequences. Option A is immediate dismissal: fast relief but high procedural, dignity, legal, precedent and trust risks. Option B is a defined supported-improvement process with clear objectives, regular reviews, adjustments where appropriate and a fair outcome: it imposes time and management cost but preserves options. Option C is role redesign or temporary redistribution: it may address root causes but may shift burdens to colleagues.
Recommended approach and safeguards. Choose Option B initially, subject to evidence and applicable law, while testing whether role or management factors contribute. Set time-bound objectives, support, a named reviewer, employee voice, confidentiality limits and review points. If improvement does not occur despite fair support, formal action may become more defensible.
Review questions. Did performance improve? Was support accessible? Did team workload reduce? Were comparable cases treated consistently? Did the employee understand the process? What system factors were learned?
Case 2 — AI recruitment disadvantaging a demographic group
Facts and uncertainties. An AI screening tool reduces administrative time but appears to reject a higher proportion of applicants from a demographic group. It is unclear whether the difference reflects job-related criteria, biased historical data, proxy variables, unequal application patterns, accessibility barriers or measurement error.
Stakeholders and values. Candidates, the demographic group, recruiters, hiring managers, the organisation, customers, regulators and society are affected. Efficiency, inclusion, validity, privacy, transparency, accountability and fairness conflict.
Frameworks and evidence. Consequentialism weighs efficiency against exclusion, lost talent and trust. Duty ethics requires non-discrimination, honest communication and human responsibility. Virtue ethics asks whether a trustworthy employer would deploy an opaque tool. Justice examines selection rates, opportunity, process voice and correction. Obtain validation data, subgroup outcomes, job-analysis evidence, accessibility testing, vendor documentation, human-review records and candidate feedback.
Legal and professional considerations. Verify current equality, data-protection and automated-decision rules. The UK Equality Act 2010 provides workplace protection from discrimination.[6] Do not assume that a vendor’s “bias-free” claim proves fairness.
Options and consequences. Option A deploy unchanged: efficient but ethically weak if adverse impact is unexamined. Option B pause and audit: slower but improves evidence and accountability. Option C use the tool only for low-risk administrative support with structured human review: preserves some efficiency while limiting harm.
Recommended approach and safeguards. Choose B followed, if justified, by C. Establish thresholds, subgroup monitoring, independent validation, human override, candidate communication, correction routes and a stop rule.
Review questions. Are disparities reducing? Are decisions valid for job performance? Are reviewers overriding systematically? Can rejected candidates understand and challenge material errors? Is the model still fit for purpose after job or labour-market changes?
Case 3 — Employee monitoring
Facts and uncertainties. Productivity has fallen and leadership proposes recording computer activity. The actual performance problem, data purpose, proportionality, alternatives, employee expectations and impact on different roles are uncertain.
Stakeholders and values. Employees, managers, customers, IT, HR, representatives and leaders are affected. Privacy, trust, efficiency, autonomy, accountability and security conflict.
Frameworks and evidence. Compare aggregated outcomes, output measures, workload data, process bottlenecks and employee voice before collecting intrusive data. The ICO states that monitoring worker data interacts with data-protection rights and that its guidance is under review due to legislative change.[5]
Options and consequences. Options are no monitoring with management intervention; limited transparent monitoring for a defined security purpose; or continuous individual surveillance. The third may produce apparent control but also stress, gaming, mistrust and privacy harm.
Recommended approach and safeguards. Define the problem first; prefer the least intrusive effective measure; consult employees; provide notice, purpose limitation, access controls, retention limits, human review and an appeal route. Do not use productivity proxies as automatic performance judgements without validation.
Review questions. Did the intervention solve the defined problem? What privacy or wellbeing effects occurred? Did behaviour become more productive or merely more visible? Were some roles unfairly measured?
Case 4 — Redundancy with disproportionate effect
Facts and uncertainties. Cost reduction is necessary, but proposed selection criteria disproportionately affect one group. The business case, alternatives, data quality, consultation and criteria validity require testing.
Stakeholders and values. Employees, representatives, managers, customers, owners and communities are affected. Sustainability, equality, fairness, compassion, transparency and efficiency conflict.
Frameworks and evidence. Assess financial necessity, alternative savings, selection validity, equality impact, consultation evidence, redeployment, training, voluntary options and long-term capability. Justice is central but should be tested against consequences and duties.
Recommended approach and safeguards. Pause final selection until the disproportionate effect is understood and alternatives are assessed. Use objective, job-related and consistently applied criteria; consult genuinely; offer support and challenge; document reasons; monitor outcomes. Legal consultation and equality requirements must be checked for the applicable jurisdiction.
Review questions. Were alternatives examined? Was the burden avoidably concentrated? Did consultation change the proposal? What will the organisation learn about workforce planning?
Case 5 — Pay transparency
Facts and uncertainties. Managers oppose transparency because they expect dissatisfaction. It is uncertain whether pay differences are justified, whether criteria are understood, whether data is accurate and whether employees can challenge errors.
Stakeholders and values. Employees, managers, leaders, candidates, representatives and owners are affected. Transparency, privacy, fairness, trust, autonomy and managerial flexibility conflict.
Frameworks and evidence. Audit pay by role and relevant characteristics; assess job architecture, progression criteria, market data, employee views and the consequences of different transparency levels. Do not equate silence with harmony.
Recommended approach and safeguards. Use graduated transparency: publish clear pay principles and ranges, explain progression, audit gaps, protect individual privacy and create correction routes. The ethical duty is not to publish every individual salary, but to make the system intelligible and defensible.
Review questions. Can employees explain how pay is set? Are gaps narrowing where unjustified? Has transparency improved trust or merely exposed unresolved inconsistency?
Case 6 — High performer misconduct
Facts and uncertainties. A commercially critical employee breaches policy. The seriousness, intent, impact, evidence, precedent and alternatives must be established.
Stakeholders and values. The employee, affected colleagues, customers, managers, HR, leaders and the organisation are affected. Accountability, consistency, loyalty, commercial continuity, safety and justice conflict.
Frameworks and evidence. A consequentialist may consider loss of revenue; duty ethics rejects special exemption from core duties; virtue ethics asks whether leaders model integrity; justice asks whether status alters the process. Verify facts and comparator cases.
Recommended approach and safeguards. Apply a fair, independent and proportionate process. Commercial importance may affect continuity planning, not whether misconduct is ignored. Protect evidence, prevent retaliation and explain the principle that performance does not purchase immunity.
Review questions. Was the person treated consistently? Did leaders model the standard? Did the organisation address incentives that encouraged misconduct?
Case 7 — Whistleblowing involving a senior executive
Facts and uncertainties. An employee reports serious concerns involving a powerful senior executive. The allegation, evidence, reporting route, retaliation risk and independence of investigation are uncertain.
Stakeholders and values. The reporter, executive, witnesses, board, employees, customers, regulators and public may be affected. Confidentiality, due process, truth, safety, accountability and reputation conflict.
Frameworks and evidence. Protect the reporter while respecting the subject’s right to a fair investigation. Use independent investigators, preserve evidence, avoid conflicts, define escalation and distinguish allegation from finding.
Recommended approach and safeguards. Escalate outside the executive’s control, use a protected channel, document decisions, assess interim risk, communicate appropriately and prohibit retaliation. Do not promise absolute confidentiality if disclosure is legally or ethically required.
Review questions. Was the investigation independent? Was the reporter protected? Were findings acted upon regardless of status? Were systemic causes identified?
Case 8 — Wellbeing versus performance targets
Facts and uncertainties. The organisation wants to improve wellbeing while maintaining demanding targets. It is unclear whether targets are achievable, whether wellbeing initiatives change work design, and whether managers are rewarded for sustainable performance or excessive output.
Stakeholders and values. Employees, managers, customers, leaders and owners are affected. Wellbeing, productivity, autonomy, sustainability, care and accountability conflict.
Frameworks and evidence. Examine workload, hours, absence, turnover, quality, psychological safety, employee voice and target-setting. A wellbeing app cannot ethically compensate for an impossible workload.
Recommended approach and safeguards. Redesign work and targets, train managers, create protected recovery time, monitor unintended effects and include employee voice. Retain accountability for performance while refusing to treat foreseeable harm as a private employee problem.
Review questions. Are workload and wellbeing indicators improving together? Do incentives reward sustainable behaviour? Can employees raise concerns without career penalty?
Ai and ethics in hr
AI is not a single ethical issue. Its risks depend on purpose, data, decision significance, affected groups, human oversight, vendor arrangements and the organisational system around it.
| Application | Ethical risks | Minimum analytical questions |
|---|---|---|
| Recruitment | Proxy discrimination, opacity, privacy, exclusion | Is it valid, necessary, accessible, monitored and challengeable? |
| Performance management | Measurement error, gaming, surveillance, automation bias | What is being measured, and does it represent performance fairly? |
| People analytics | Re-identification, speculative inference, function creep | Is the purpose legitimate, proportionate and understood? |
| Monitoring | Privacy intrusion, stress, unequal effects | Is there a less intrusive effective alternative? |
| Workforce planning | Stereotyping, displacement, opaque forecasts | Who is affected by predictions and who can challenge them? |
| Generative AI | Confidentiality, hallucination, bias, deskilling | What data may be entered, who checks outputs and who remains accountable? |
Ethical AI in HR requires fairness, transparency, explainability, privacy, accountability, bias mitigation, human oversight and proportionality. “Human in the loop” is not enough if the human lacks time, expertise or authority to disagree. Human oversight must be meaningful.
Connect this section to Recruitment Technology in 2026: Opportunity or Risk? and verify current regulatory sources before publication. Keep claims current because AI governance and data rules are changing.
Ethics in performance management
Performance management creates dilemmas around ratings, dismissal, capability procedures, manager discretion, performance-improvement plans and continuous feedback. The ethical question is not whether performance matters; it is whether the system measures relevant contribution, gives people a fair chance to succeed, treats context responsibly and avoids converting managerial preference into objective truth.
A defensible process uses clear expectations, valid evidence, timely feedback, employee voice, appropriate support, consistency with room for relevant differences, independent review and proportionate consequences. It should also inspect system causes: poor role design, inadequate resources, conflicting objectives, weak management, inaccessible technology or a culture that rewards overwork.
Connect to Annual Appraisal vs Continuous Feedback: What Does the Evidence Say? and Line Managers as the Critical Variable in Performance Management. The internal link should be inserted only after the publisher verifies its URL.
Ethics in reward
Reward dilemmas arise in performance-related pay, executive compensation, pay inequality, transparency, recognition and incentives. Incentives can focus effort, but they can also encourage gaming, rivalry, unsafe conduct, short-termism or the neglect of unmeasured work such as care, mentoring and inclusion.
Use the tool to ask whether the measure is valid, whether the distribution is fair, whether people can influence the result, whether risk is shifted downward, whether vulnerable groups are disadvantaged and whether the reward system expresses the organisation’s stated values. A high reward may be legal and market-aligned while still being difficult to justify if it is disconnected from stewardship, harm or collective contribution.
Connect to Performance-Related Pay: Effective Motivator or Management Myth?
Ethics in resourcing
Selection, recruitment algorithms, referrals, diversity initiatives, employer branding, candidate privacy and background checks all involve competing values. Efficiency may conflict with equal opportunity; employer branding may conflict with honesty; referrals may improve trust while reproducing homogeneity; background checks may protect safety while intruding into privacy or relying on irrelevant history.
Ethical resourcing uses job-related criteria, accessible processes, reliable evidence, proportionate information collection, clear candidate communication and routes to correct error. It does not assume that a process is fair because it is standardised or automated.
Connect to Recruitment Technology in 2026: Opportunity or Risk?
Ethics in workforce planning
Redundancies, restructuring, automation, skills-based planning, outsourcing, offshoring and contingent labour raise questions about who bears risk and how the organisation defines responsibility. A workforce plan may be financially rational yet ethically weak if it treats people as cost units, ignores community impact or assumes displaced workers can easily transition.
Analyse alternatives, time horizons, capability, consultation, equality effects, supply-chain responsibility, skills investment and the reversibility of the decision. Ethical planning includes honest communication about uncertainty rather than presenting forecasts as destiny.
Connect to Workforce Planning Maturity: Where Most Organisations Really Are and assess the claims against current evidence.
High road vs low road
A low-road approach typically emphasises cost minimisation, labour flexibility, short-term efficiency and limited employee investment. A high-road approach typically emphasises development, participation, capability, sustainable performance and longer-term value.
Neither label is automatically ethical or commercially superior. A high-road policy can become symbolic or financially unsustainable; a low-cost intervention can protect jobs or customers in a crisis. Compare costs, feasibility, trade-offs, context, stakeholder impact, implementation quality and time horizon.
| Question | Low-road risk | High-road risk |
|---|---|---|
| Who bears cost? | Employees and suppliers may absorb volatility. | Investment may be unavailable or diverted from urgent needs. |
| What is measured? | Immediate cost and output. | Long-term capability and quality, which may be harder to attribute. |
| What can go wrong? | Turnover, distrust, low skill and hidden harm. | Overpromising, slow decisions or inequitable access to development. |
| Ethical test | Is cost minimisation necessary and fairly distributed? | Is investment realistic, inclusive and genuinely connected to outcomes? |
Connect to High Road vs Low Road: A Strategic Guide.
Ethical decision-making as a system
A decision changes behaviour, behaviour changes managerial practice, managerial practice shapes culture, culture becomes embedded in HR systems, and systems influence organisational outcomes.
Decision → employee behaviour → manager behaviour → culture → HR systems → organisational outcomes → feedback into future decisions.
For example, exempting a high performer from misconduct consequences may increase short-term revenue while teaching others that status outweighs standards. A monitoring system may increase recorded activity while reducing trust and innovation. A redundancy process may save cost while losing critical capability and creating survivor anxiety.
Systems thinking requires a second-order review: What behaviours will this decision encourage? What will managers learn they can do? Which groups will adapt at greatest cost? What will become normal? Connect this analysis to Systemic Thinking in HRM: Service-Profit Chain, Bath Model, Balanced Scorecard.
The level 7 approach
Level 7 analysis should move beyond description and assertion.
| Move | Question | Evidence of quality |
|---|---|---|
| Describe | What is happening? | Accurate, neutral account of facts and uncertainty. |
| Analyse | Why is it ethically problematic? | Explicit conflict among values, interests, duties or rights. |
| Apply | Which frameworks are relevant? | Frameworks are used to interpret the case, not merely named. |
| Evaluate | What are the strengths and weaknesses of each option? | Consequences, duties, fairness, feasibility and risk are compared. |
| Challenge | What assumptions, power relationships and unintended effects exist? | Alternative explanations and under-heard voices are considered. |
| Judge | Which option is most ethically defensible and why? | A reasoned conclusion is stronger than alternatives. |
| Recommend | How should it be implemented? | Safeguards, ownership, communication and review are specified. |
Command word guide
| Command word | What the tool requires |
|---|---|
| Explain | Define the dilemma and relevant principles clearly. |
| Analyse | Examine relationships between stakeholders, values, evidence and consequences. |
| Discuss | Present competing ethical perspectives and tensions. |
| Evaluate | Assess alternative options using criteria and evidence. |
| Critically evaluate | Challenge assumptions, evidence, frameworks and contextual relevance. |
| Recommend | Select and justify the most defensible course of action. |
| Reflect | Examine personal assumptions, judgement, learning and future practice. |
Command word guide
When answering “explain,” do not simply state that a decision is unfair; define what fairness means and identify the relevant process or outcome. When asked to analyse, show relationships: how a manager’s power affects the evidence, how evidence affects the stakeholder assessment and how values create tension. When asked to critically evaluate, question whether the evidence is valid, whether the framework fits the context, whether alternatives were considered and whether the recommendation creates unintended harm. When asked to recommend, make a decision and explain implementation, safeguards and review rather than ending with “it depends.”
Model level 7 paragraph
Claim. Immediate dismissal may protect short-term team output, but it is not automatically the most ethically defensible response to underperformance. Ethical framework. A consequentialist analysis recognises potential benefits for colleagues and customers, while a justice-based analysis asks whether the employee received a fair process and whether relevant circumstances were considered. Evidence. The recommendation should therefore be informed by objective performance records, workload data, the employee’s account, support offered and comparator cases rather than managerial perception alone. Application. If expectations were unclear and support was limited, a defined supported-improvement process may reduce preventable harm while preserving accountability. Counterargument. The team should not be required to absorb indefinite underperformance, and compassion cannot become an exemption from reasonable standards. Critical evaluation. This counterargument is persuasive if the process includes time limits, measurable objectives and consequences, but weaker if “consistency” is used to ignore materially different circumstances or if dismissal is treated as inevitable before evidence is reviewed. Professional judgement. The most defensible initial recommendation is therefore a time-bound, evidence-informed capability process with appropriate support, meaningful employee voice, confidentiality safeguards and a review point at which formal action remains available if improvement does not occur.
This demonstrates Level 7 thinking because it moves from claim to theory, evidence, application, counterargument, critical evaluation and judgement. It does not confuse empathy with analysis, list theories without using them or hide uncertainty behind a confident conclusion.
Weak vs strong ethical analysis
| Topic | Weak analysis | Stronger analysis |
|---|---|---|
| Performance | “It would be unfair to dismiss the employee.” | “Dismissal may protect team performance, but its defensibility depends on clear expectations, reliable evidence, consistent process and consideration of relevant circumstances.” |
| AI recruitment | “AI is biased.” | “The observed demographic disparity is an ethical warning, not proof of cause; the organisation should test validity, proxies, data quality, accessibility and human review before deployment.” |
| Monitoring | “Employees deserve privacy.” | “Privacy concerns require a necessity and proportionality test: the organisation should identify the objective, compare less intrusive alternatives and govern access, notice, retention and challenge.” |
| Redundancy | “Redundancy is unethical.” | “Cost reduction may be necessary, but selection is ethically weaker where criteria are weakly job-related, consultation is tokenistic or burdens fall disproportionately without justification.” |
| Pay | “Transparency will solve pay inequality.” | “Transparency can expose unjustified differences, but it must be paired with valid job architecture, privacy safeguards, correction routes and accountable progression criteria.” |
| Whistleblowing | “The whistleblower must be believed.” | “The report should be taken seriously and the reporter protected, while the subject retains a fair investigation and findings are based on independent evidence.” |
| Diversity | “Treating everyone identically is fair.” | “Equality of treatment can reproduce disadvantage where relevant barriers differ; fairness requires justified criteria, accessibility and monitoring of outcomes.” |
| Wellbeing | “Wellbeing is more important than performance.” | “Sustainable performance requires attention to wellbeing, but the ethical task is to design workable targets and support rather than replace accountability with sentiment.” |
| Confidentiality | “HR must keep everything secret.” | “Confidentiality protects privacy and process integrity, but appropriate transparency is required for people to understand, challenge and trust the decision.” |
| Employee voice | “We consulted, so the decision is ethical.” | “Consultation improves information and legitimacy only when participation is safe, representative, informed and capable of influencing the outcome.” |
| Discipline | “Rules must always be applied identically.” | “Consistency matters, but justice requires comparable cases to be compared and relevant differences to be assessed rather than ignored.” |
| Promotion | “The best performer deserves promotion.” | “Promotion criteria should be valid, transparent and inclusive; performance in one role does not automatically demonstrate leadership capability or justify preference.” |
| Flexible working | “Everyone should work the same way.” | “A common standard may support coordination, but fairness requires examining role requirements, accessibility, caring responsibilities and evidence of actual impact.” |
| Restructuring | “The cheapest structure is the best.” | “Cost is relevant, but ethical evaluation also considers capability, transition harm, consultation, equality effects, service quality and long-term sustainability.” |
| High performers | “We cannot afford to lose the star.” | “Retention risk matters, but exempting misconduct from fair process can damage trust, justice and future behaviour more than the immediate loss.” |
Common student mistakes
- Treating ethics as personal opinion rather than reasoned judgement.
- Assuming that legal compliance settles the ethical question.
- Listing ethical theories without applying them to the facts.
- Applying only one framework and ignoring competing insights.
- Naming stakeholders without analysing their interests, rights or power.
- Ignoring who defines the problem and controls the evidence.
- Making unsupported claims about impact, bias or motivation.
- Treating fairness as identical treatment.
- Presenting assumptions as established facts.
- Jumping directly to a recommendation without defining the decision.
- Ignoring realistic alternatives and mitigations.
- Considering only intended, short-term consequences.
- Ignoring context and treating a policy as self-executing.
- Treating professional codes as complete answers.
- Assuming the most profitable option is ethically best.
- Assuming employee interests always override organisational interests.
- Assuming organisational interests always override employee interests.
- Confusing empathy with ethical reasoning.
- Ignoring implementation, communication and review.
- Failing to explain why the preferred option is stronger than alternatives.
- Treating employee consultation as proof of legitimacy.
- Treating AI output or dashboards as neutral evidence.
- Using confidentiality to avoid accountability.
- Using reputational risk as a substitute for ethical reasoning.
- Writing “it depends” without identifying what it depends on or what decision follows.
One-page checklist
Before deciding
- [ ] What exactly is the decision?
- [ ] What are the established facts?
- [ ] What is uncertain, contested or unverified?
- [ ] Who is affected, vulnerable, absent or under-heard?
- [ ] What values conflict?
- [ ] What rights and duties exist?
- [ ] What evidence is available, and how strong is it?
- [ ] What evidence challenges the preferred option?
- [ ] What realistic options exist, including less harmful alternatives?
- [ ] What does each ethical framework suggest?
- [ ] What are the short-term and long-term consequences?
- [ ] What unintended consequences could occur?
- [ ] Is the process fair, respectful and explainable?
- [ ] Is the decision necessary, effective and proportionate?
- [ ] What professional standards apply?
- [ ] What current legal requirements apply?
- [ ] Can affected people challenge or correct the decision?
- [ ] Who is accountable for implementation?
- [ ] What safeguards reduce foreseeable harm?
- [ ] When and how will the decision be reviewed?
The final test
Could I explain this decision transparently to the affected employee, the wider workforce, a professional body and senior leadership? Could I justify it with evidence, defend the process as fair, and make the same decision if the identities of the people involved were reversed?
Reputational discomfort alone does not determine ethics, but it can reveal hidden assumptions and prompt a useful transparency test.
Editable ethical dilemma worksheet
Ethical dilemma
Decision required
Facts and evidence status
Uncertainties, assumptions and alternative explanations
Stakeholders, power and voice
Competing values
Rights, duties, law, policy and professional standards
Evidence needed and evidence quality
Option A: description, benefits, harms and safeguards
Option B: description, benefits, harms and safeguards
Option C: description, benefits, harms and safeguards
Ethical options matrix
| Option | Benefits | Harms | Rights/duties | Fairness | Stakeholder impact | Practicality | Ethical risks |
|---|---|---|---|---|---|---|---|
| A | |||||||
| B | |||||||
| C |
Utilitarian/consequentialist analysis
Deontological/duty analysis
Virtue ethics analysis
Justice/fairness analysis
Proportionality test
Necessity: _______
Effectiveness: _______
Least harmful alternative: _____
Fairness: ________
Transparency: ______
Accountability: ______
Review: ___________
Preferred option
Why is it more defensible than the alternatives?
Implementation safeguards and communication plan
Review measures, trigger points and learning
Self-assessment
Rate each statement from 1 = not yet confident to 5 = highly confident.
| Statement | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| I can distinguish an ethical dilemma from a general problem. | |||||
| I can separate facts from assumptions. | |||||
| I can identify competing stakeholder interests. | |||||
| I can identify competing values. | |||||
| I understand consequentialism. | |||||
| I understand deontology. | |||||
| I understand virtue ethics. | |||||
| I understand justice and fairness approaches. | |||||
| I can compare ethical frameworks rather than merely name them. | |||||
| I can identify power dynamics. | |||||
| I can integrate evidence. | |||||
| I can distinguish legal from ethical considerations. | |||||
| I can assess proportionality. | |||||
| I can make a defensible professional judgement. | |||||
| I can explain why my preferred option is stronger than alternatives. |
This is an educational self-assessment and is not a validated CIPD assessment instrument.
FAQ
1. What is an ethical dilemma in HR?
It is a decision involving competing legitimate values, duties, rights, interests or stakeholder claims where the best course is not immediately obvious.
2. What is the difference between an ethical issue and an ethical dilemma?
An ethical issue is a morally significant feature of a situation. An ethical dilemma exists when significant ethical considerations conflict and must be balanced or prioritised.
3. Is an ethical dilemma always a conflict between right and wrong?
No. Many HR dilemmas are conflicts between two defensible goods, such as compassion and accountability, privacy and transparency, or efficiency and inclusion.
4. What are the four main ethical frameworks?
This tool uses consequentialism, deontology or duty ethics, virtue ethics, and justice or fairness. They should be treated as complementary lenses rather than competing doctrines where one always wins.
5. Which ethical framework is best?
None is universally best. The strongest analysis explains what each reveals, where each is limited and why a particular combination is appropriate to the context.
6. Can something be legal but unethical?
Yes. Legal compliance may be necessary but can leave questions about dignity, proportionality, power, transparency, fairness and preventable harm unresolved.
7. How should HR balance employee and organisational interests?
Do not assume that one side automatically prevails. Identify legitimate interests, rights, duties, power, evidence and consequences, then choose the option that is proportionate and most defensible.
8. What role does evidence play in ethical decisions?
Evidence tests assumptions, estimates impact and identifies alternatives. It does not replace ethical judgement because data cannot by itself decide what people are owed.
9. How does the CIPD Profession Map relate to ethics?
It expresses professional expectations such as ethical practice, evidence-based decision-making, courage, inclusion and valuing people. It guides professional conduct but does not mechanically resolve every dilemma.[1]
10. How should Level 7 students analyse ethical dilemmas?
They should identify, analyse, apply, evaluate, challenge, judge and recommend. The work should integrate theory, evidence, power, context, alternatives and implementation.
11. How important is stakeholder analysis?
It is essential because HR decisions distribute benefits and burdens. A stakeholder map should include voice, vulnerability, rights, interests and power, not just names.
12. Why does power matter in HR ethics?
Power affects who defines the problem, controls information, speaks safely, challenges decisions and bears consequences. Ignoring it can make a formally neutral process ethically misleading.
13. How should HR approach AI-related ethical dilemmas?
Test necessity, validity, bias, transparency, privacy, explainability, accessibility, human oversight, accountability and proportionality. Maintain a meaningful route to challenge and correct errors.
14. How can fairness be assessed?
Assess distributive, procedural, interactional and informational fairness. Ask whether relevant differences were considered and whether the decision can be justified consistently.
15. What is the biggest mistake students make when analysing ethics?
The most common error is asserting a conclusion—“this is unfair”—without defining the standard of fairness, applying frameworks, weighing alternatives, using evidence and explaining why the recommendation is stronger.
16. Is empathy enough for ethical HR practice?
No. Empathy can reveal experience and vulnerability, but ethical reasoning also requires evidence, duties, fairness, consequences, accountability and professional judgement.
17. Does consultation make a decision ethical?
No. Consultation may be tokenistic or unsafe. Its ethical value depends on representation, information, influence, feedback and protection from retaliation.
18. How much confidential information should HR disclose?
Enough to provide a truthful and meaningful explanation, but not more than is legitimate and proportionate. Protect personal data and investigation integrity while avoiding secrecy used as an accountability shield.
References
The following references use a Harvard-style presentation. Web sources should be checked for currency at publication and accessed on 15 August 2026. The numbered links support the inline citations used in this resource.
[1] Chartered Institute of Personnel and Development (CIPD) (n.d.) The Profession Map: Core behaviours and Ethical practice. Available at: CIPD Profession Map core behaviours and CIPD ethical practice (Accessed: 15 August 2026).
[2] Chartered Institute of Personnel and Development (CIPD) (2023) Code of Conduct and Ethics. Available at: CIPD Code of Conduct and Ethics (Accessed: 15 August 2026).
[3] Chartered Institute of Personnel and Development (CIPD) (2025) Evidence-based HR: Make better decisions and step up your influence. Available at: CIPD evidence-based profession guide (Accessed: 15 August 2026).
[4] Chartered Institute of Personnel and Development (CIPD) (n.d.) Ethical practice and the role of people professionals. Available at: CIPD ethics factsheet (Accessed: 15 August 2026).
[5] Information Commissioner’s Office (ICO) (n.d.) Employment practices and data protection: monitoring workers. Available at: ICO monitoring workers guidance (Accessed: 15 August 2026).
[6] UK Government (2026) Equality Act 2010: guidance. Available at: GOV.UK Equality Act guidance (Accessed: 15 August 2026).
[7] Markkula Center for Applied Ethics, Santa Clara University (n.d.) A Framework for Ethical Decision Making. Available at: Santa Clara University ethics framework (Accessed: 15 August 2026).
[8] Freeman, R.E. (1984) Strategic Management: A Stakeholder Approach. Boston: Pitman. Use a library or publisher record to verify the edition used.
[9] Greenberg, J. (1990) ‘Organizational justice: Yesterday, today, and tomorrow’, Journal of Management, 16(2), pp. 399–432. Verify the DOI and pagination against the library database used by the publisher.
[10] Rawls, J. (1971) A Theory of Justice. Cambridge, MA: Harvard University Press. Use the edition cited by the institution or assignment.
[11] Aristotle (2009) The Nicomachean Ethics, translated by W.D. Ross and revised by L. Brown. Oxford: Oxford University Press. Use the edition required by the course.
[12] Kant, I. (1785/2012) Groundwork of the Metaphysics of Morals, translated and edited by M. Gregor and J. Timmermann. Cambridge: Cambridge University Press. Use the edition required by the course.
[13] Mill, J.S. (1863/2001) Utilitarianism. Indianapolis: Hackett. Use the edition required by the course.
[14] Treviño, L.K. (1986) ‘Ethical decision making in organizations: A person-situation interactionist model’, Academy of Management Review, 11(3), pp. 601–617. Verify the DOI and pagination against the library database used by the publisher.
[15] Rest, J.R. (1986) Moral Development: Advances in Research and Theory. New York: Praeger.
[16] Bies, R.J. and Moag, J.S. (1986) ‘Interactional justice: Communication criteria of fairness’, in Lewicki, R.J., Sheppard, B.H. and Bazerman, M.H. (eds.) Research on Negotiation in Organizations. Greenwich, CT: JAI Press, pp. 43–55.
[17] Nishii, L.H. (2013) ‘The benefits of climate for inclusion for gender-diverse groups’, Academy of Management Journal, 56(6), pp. 1754–1774. Verify the DOI and pagination before publication.
[18] International Labour Organization (ILO) (2019) Safety and Health at the Heart of the Future of Work: Building on 100 Years of Experience. Geneva: ILO. Verify the current publication record if cited in a final academic assignment.
[19] European Commission (2024) Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). Available through the official EUR-Lex database. Verify application dates and relevance to the organisation’s jurisdiction before use.
[20] ACAS (n.d.) Whistleblowing at work. Available at: ACAS whistleblowing guidance (Accessed: 15
The Level 7 Rule
Do not ask only, “What is the right answer?” Ask, “Which option can I justify most convincingly when the evidence, values, rights, consequences, stakeholders and professional responsibilities are all considered?”
