03 · What You Need to Know
How to Make a Decision When Every Option Has an Ethical Cost
First, Distinguish a Genuine Ethical Dilemma From an Inconvenient Choice
Researchers sometimes describe a decision as an ethical dilemma when one option is simply more expensive, slower, administratively difficult, or methodologically inconvenient.
Those constraints matter, but they do not automatically create competing ethical claims.
A genuine ethical dilemma arises when available options protect and sacrifice ethically important interests. Protecting confidentiality may conflict with preventing serious harm. Respecting autonomous choice may conflict with concern about unacceptable risk. Protecting a population through exclusion may conflict with justice and the need for evidence relevant to that population.
Before declaring that there is “no ethical option,” state exactly what ethical value is compromised by each realistic alternative.
Some Ethical Problems Really Do Resist Clean Resolution
The Belmont Report was not written on the assumption that ethical principles always produce one indisputable answer. It explicitly states that its principles cannot always resolve particular ethical problems beyond dispute. It also gives examples in which different claims within the same principle can conflict.
That matters because ethical ambiguity is not necessarily evidence that the researcher has reasoned badly. Some situations contain genuinely competing obligations.
The mistake is not encountering an ethical remainder. The mistake is pretending that one does not exist because an ethics application looks more reassuring when every box ends in “risk adequately addressed.”
Do Not Begin by Asking Which Principle Wins
A common response to a difficult dilemma is to search for a hierarchy: autonomy first, perhaps, or preventing harm above everything else.
Belmont provides no such universal ranking. Historical reflections from people involved in the Commission's work confirm that the Commission deliberately declined to make one of its three principles automatically take precedence in hard cases. Which consideration should carry greater weight depends on the facts and what is at stake.
That does not mean every principle always deserves equal weight. A serious and probable harm can carry more weight than a minor inconvenience. A substantial violation of autonomy can outweigh a speculative benefit.
The point is that the relative weight has to be justified in the case rather than imported from a universal ranking.
Get the Facts as Clear as Possible Before Making the Ethical Judgment
Many apparent ethical dilemmas become easier once uncertain facts are clarified.
How probable is the feared harm? How severe would it be? Is the participant actually identifiable? Is there genuinely a legal duty to disclose? Is the participant's decision voluntary? Is the risky procedure scientifically necessary? Would stopping the study create additional risk? Can another method answer the same question?
Belmont recommends a systematic, nonarbitrary approach to risk-benefit assessment, including careful accumulation of relevant information, examination of the validity of research assumptions, attention to the nature, probability, and magnitude of risks, and systematic consideration of alternatives.
Ethical reasoning cannot eliminate uncertainty, but it should avoid manufacturing ethical conflict from factual assumptions that were never checked.
Separate What You Know From What You Are Predicting
Difficult ethical decisions often involve uncertain futures.
A participant might experience harm. Disclosure might prevent something serious. Continuing research might generate important knowledge. Exclusion might perpetuate an evidence gap.
Researchers should therefore distinguish evidence from prediction and avoid giving uncertain outcomes the rhetorical certainty needed to justify a preferred choice.
A highly speculative catastrophic harm should not automatically outweigh a likely substantial infringement merely because the imagined consequence is dramatic. Likewise, an uncertain future social benefit should not be described as though it were guaranteed.
Identify Who Bears the Ethical Cost
An option can look reasonable when costs and benefits are described in aggregate while becoming much less comfortable when researchers ask who receives each one.
Does one participant bear substantial risk so that future populations might benefit? Does a marginalized community provide the data while another population receives the intervention? Does protecting the institution's reputation shift a burden onto participants? Does preserving confidentiality protect one person while exposing another to serious risk?
Justice matters because ethical costs are not interchangeable simply because they can all be placed in the same paragraph.
Look for an Option That Changes the Structure of the Dilemma
Before choosing between two ethically costly options, ask whether the research can be redesigned so that less must be sacrificed.
Could identifiable data be separated from research responses? Could an independent recruiter reduce pressure? Could a risky procedure be replaced with a less invasive measure? Could confidentiality limits be defined prospectively? Could a vulnerable population be included with additional safeguards rather than excluded completely? Could the research question be narrowed?
This is the practical extension of reasoning through conflicting ethical principles: do not balance competing harms until you have seriously tested whether the conflict itself can be reduced.
Some Ethical Costs Can Be Minimized Even When They Cannot Be Eliminated
A dilemma does not necessarily disappear when a safeguard is introduced. The safeguard may simply make one option more defensible.
For example, research involving sensitive disclosures may never guarantee absolute confidentiality. Researchers may nevertheless minimize identifiable data, restrict access, establish secure storage, define disclosure procedures, communicate genuine limits accurately, and avoid collecting information they do not need.
The remaining risk still exists, but it may become proportionate to the research purpose.
Ethical decision-making often concerns reducing avoidable moral cost rather than eliminating every conceivable risk.
Consider Whether the Ethical Cost Is Necessary
A particularly important question is whether the compromised interest must actually be compromised to achieve a legitimate research objective.
If a privacy intrusion produces no information necessary to the study, it is difficult to justify. If deception adds nothing scientifically important, the autonomy cost loses much of its rationale. If a high-risk procedure can be replaced by a lower-risk procedure that answers the question adequately, the additional risk requires explanation.
Necessity does not automatically make an ethical cost acceptable. But lack of necessity often makes it much harder to defend.
Compare the Magnitude and Probability of the Remaining Ethical Costs
Once avoidable problems have been removed, compare what remains.
Belmont cautions that balancing risks and benefits is rarely a precise quantitative exercise. Nevertheless, it calls for systematic analysis rather than intuition alone. Researchers should distinguish the nature, probability, and magnitude of possible harm as clearly as possible.
Useful considerations include severity, probability, duration, reversibility, number of people affected, distribution of burden, certainty of anticipated benefit, and whether important rights or interests are being infringed.
This is not an ethics calculator. A spreadsheet cannot tell you that 7.4 units of beneficence outweigh 6.9 units of autonomy. The structure simply makes the reasoning less arbitrary.
Some Ethical Constraints Should Not Be Treated as Ordinary Trade-Offs
Not every ethical consideration can be converted into a cost that becomes acceptable once enough benefit is placed on the other side.
Belmont states that brutal or inhumane treatment of human participants is never morally justified. Its discussion of risk-benefit assessment also emphasizes that participant rights must be protected even when benefits beyond the individual are considered.
This matters because “there is no perfect option” should not become a rhetorical device for rationalizing conduct that crosses a substantive ethical boundary.
Some options should be removed from consideration rather than balanced against expected scientific gain.
Applicable Law and Institutional Requirements Still Constrain the Decision
A researcher may regard several options as ethically arguable while only some are legally or institutionally permitted.
Ethical reasoning does not normally authorize unilateral violation of applicable law, regulation, or institutional requirements. Apparent conflicts should be taken to the appropriate authority, and lawful alternatives, amendments, exceptions, or formal reconsideration should be explored where available.
The feasible ethical choice therefore has to be distinguished from an option that the researcher has no authority to implement.
Independent Consultation Is Particularly Important in Hard Cases
Researchers are most likely to need outside judgment precisely when they are most invested in the outcome.
Consultation might involve the ethics committee, supervisor, research ethics office, participant or community representatives, legal advisers, data-protection specialists, clinicians, safeguarding professionals, methodological experts, or other relevant authorities.
The appropriate person depends on the dilemma.
Consultation also creates a check against motivated reasoning. A researcher who desperately needs the study completed before graduation, funding expiry, or publication submission is not suddenly disqualified from ethical thought, but recognizing the incentive is healthy.
Do Not Shop for the Answer You Want
Seeking a second perspective is different from repeatedly asking people until someone endorses the preferred option.
Good consultation presents the difficult facts honestly, including information that weakens the researcher's preferred position. It seeks scrutiny rather than permission.
If several informed reviewers disagree, document the disagreement and determine which institutional or regulatory decision governs the research. Ethical disagreement can remain genuine even when an authorized decision must ultimately be made.
Temporary Delay Can Be an Ethical Option
Researchers sometimes assume the available choices are “continue” or “terminate.”
Pausing may create room to gather missing information, consult the ethics committee, investigate a safety signal, modify consent materials, strengthen safeguards, or determine whether an apparent legal obligation actually applies.
Delay itself can have costs, especially in urgent research. But where the ethical problem arises from uncertainty that can realistically be reduced, a temporary pause may be more defensible than making an irreversible decision prematurely.
Stopping or Modifying Ongoing Research Can Become Necessary
Ethical assessment does not end when recruitment begins.
The 2024 Declaration of Helsinki requires continuous monitoring, assessment, and documentation of risks and burdens in medical research involving human participants. It states that when risks and burdens outweigh potential benefits, or definitive outcomes have been established, researchers must assess whether to continue, modify, or immediately stop the research.
This illustrates a broader point: previous justification does not guarantee future justification.
A study that was ethically defensible when approved can become difficult or impossible to justify when circumstances change.
Sunk Costs Are Not Ethical Reasons to Continue
Researchers may have spent years developing a project, secured funding, recruited participants, trained staff, and collected substantial data before a serious ethical problem becomes apparent.
Those investments are psychologically powerful. They do not by themselves justify continuing.
The relevant question is whether the research remains ethically defensible now. Past effort cannot convert unacceptable future risk into acceptable risk.
Research ethics is occasionally rude to project management in this respect.
Scientific Importance Does Not Require Every Question to Be Answered Experimentally
A research question can be important while the available methods for answering it remain ethically unacceptable.
Researchers may need to use observational evidence, secondary data, natural experiments, modeling, less invasive measures, narrower questions, or other approaches that provide weaker but ethically obtainable evidence.
Sometimes uncertainty must remain.
The existence of an unanswered scientific question does not create an entitlement to whatever experiment would answer it most decisively.
Document the Reasoning, Including What Remains Unresolved
Difficult decisions should leave an intellectual trail.
Record the relevant facts, uncertainties, ethical interests, alternatives considered, safeguards proposed, consultations obtained, applicable requirements, and reasons for the final decision.
Also record the residual ethical concern.
Documentation should not transform “we chose this option despite an unavoidable confidentiality cost” into “all ethical concerns were fully resolved.” Acknowledging what remains ethically uncomfortable is often evidence of better reasoning, not weaker reasoning.
A Defensible Decision Can Still Leave Moral Residue
Some choices remain regrettable even when they are justified.
A study may need to stop, disappointing participants who hoped it would continue. Confidentiality may need to be limited under circumstances participants would rather avoid. A population may need additional protections that constrain its autonomy. A scientifically valuable question may remain unanswered.
The fact that a decision is ethically defensible does not require pretending that nothing valuable was lost.
This remaining sense of ethical cost is sometimes described in moral philosophy as moral residue or moral remainder. The terminology is less important than the practical lesson: justification and regret can coexist.
Sometimes the Absence of an Acceptable Option Is Itself the Answer
After redesign, consultation, risk reduction, and careful comparison, researchers may still find that every feasible way of conducting the study violates an important ethical constraint or imposes burdens that cannot be sufficiently justified.
At that point, refusing to conduct the research is not a failure to solve the dilemma.
It may be the conclusion produced by solving it.
The final question is therefore whether the most ethical decision is not to conduct the study at all.
