Ethics
An effective decision can still be the wrong one
A measure may achieve its goal while causing harm that has not been properly considered or justified. We therefore ask not only which intervention might work but what is legally, professionally and morally defensible.
Rights, interests and consequences
Ethical judgement involves rights, authority, dignity, care, freedom and the fair distribution of benefits and burdens. These perspectives help us see where values conflict and who bears the consequences.
Before taking on an assignment, we agree how interests, confidentiality and access to findings will be handled.
The cheapest bid, the loyal supplier and a performance review
A fictional regional distributor must renew its packaging contract. Three bids come in. The cheapest is from a newcomer. The most expensive is from a supplier the managing director has known for twenty years, and whose owner is a family friend. The purchasing officer, Noor, writes a two-page note: the newcomer is 14% cheaper and meets every specification. A week later Noor’s manager says her next appraisal will “look closely at judgement and teamwork”. Meanwhile the director asks for a Decision Review of the “supplier strategy”, to be ready before the board meets.
Nobody has said anything improper. The director may sincerely believe that reliability is worth the premium. He may be right. Noor may be naive about delivery risk. Or she may be about to be punished for an accurate note. All three can be true at once, and the paperwork will look the same in every case. That is exactly why this is an ethical problem and not only a procurement problem.
Illustration · who is weighed
Four pulls on one decision
How judgement can quietly shift
It can begin with loyalty, irritation or concern about a difficult quarter. Then the feeling looks for an argument, and clever people find good ones quickly. “Reliability matters.” “She does not see the whole picture.” Then the argument finds allies, because agreeing with the person who signs your bonus is comfortable. At the end, the decision looks like a conclusion when it was a starting point.
Motives are difficult to judge. It helps to make the arguments clear before the decision is fixed and to ensure someone can raise an objection without personal repercussions.
The best case for the other side
We do not treat the opposing view as a villain. In the case above it has real force:
For the director
Price is not cost. A supplier who has never missed a delivery is worth something a spreadsheet does not show, and an unknown supplier who fails in December is expensive. Long relationships also lower monitoring costs. These are legitimate reasons. They can be tested: ask for the delivery records of the incumbent and references for the newcomer.
For the purchasing officer
A conflict of interest does not need bad intent to distort a decision. If the premium is paid because of friendship, the company is spending shareholders’ money on the director’s comfort. And if raising the issue gets her a poor appraisal, every other employee learns that noticing is dangerous.
A serious inquiry keeps both arguments alive long enough to see which one the facts support. It also asks something neither side likes: whose interests are not represented at all? Here, the customers who bear the cost and the colleagues watching how Noor is treated.
A sequence of operations
Ethics is not a feeling we consult at the end. It is a set of operations done in order, written down, by someone who can be held to them.
- Write down the factsSeparate what has been observed from what is believed. “The bid is 14% cheaper” is a fact. “She is undermining me” is an inference. Mark which is which.
- Check the rulesLaw, contract, procurement policy, promises already made. If a rule settles the matter, the matter is settled. Do not reopen it because the result is inconvenient.
- Test independence, consent, loyalty, fairness and harmCan the person deciding gain or lose personally? Did those affected agree to how information about them is used? Whom is each party owed loyalty? Who carries the cost? What could happen to whom?
- Estimate real effects and reversibilityNot the intended effects, the likely ones, including what the decision teaches people about speaking up. Then ask what can be undone, and what cannot.
- Decide explicitlyWrite the decision, who does what by when, and what would reopen it. “We will look into it” is not a decision.
- Open it to an independent objectionSomeone with no stake in the outcome reads the reasoning and has the standing to say it does not hold. Their objection is answered in writing.
Six questions you can use on Monday
Take a decision about people that is currently on the table. Answer these questions yourself, then discuss them with someone who sees things differently.
Six questions
- 1. What would I bet on, and what am I merely hoping?List the facts you would swear to. Everything else is a guess, even if it is a good one.
- 2. Which rule or promise already decides part of this?If one does, follow it and say so.
- 3. Who gains, who pays, and who chose?If the people who choose are also the people who gain, add a second reader.
- 4. Who would object if they knew everything I know, and could they say so without cost?If objecting costs a career, assume the objections are missing, not absent.
- 5. If I am wrong, what can be undone, by whom, and by when?The less reversible the move, the more evidence it deserves.
- 6. What would make me stop, and who will hold me to that?Name the signal now. Name the person now.
An exercise: design the dishonest version
Take your decision. Ask the group: what would we have to do to make this situation maximally dishonest?
In the procurement case the answers come fast. Bring the appraisal forward. Ask the purchasing officer for her view only after the director has stated his. Let the person who benefits write the criteria. Keep the delivery records out of the pack. Describe the dissent as a personality problem. Give the board one slide.
Then invert each answer into a protection. Criteria are fixed before bids are opened. The person with an interest does not score. Dissent is recorded and answered in writing. Nobody’s appraisal is touched for the duration of the decision unless an independent person has looked at it. It is a cheap way to find the weak points of your own process.
Two more cases, shorter
The favourite and the exception
A director wants a one-off exception to the promotion rules for a talented colleague she has mentored. She may be right about the talent. The rule exists so that being liked does not count as competence. The practical question is not “is she good” but “would we grant the same exception to someone we do not like, and who among the others passed over was asked?” If the answer is no, the exception is a favour, and should be named as one.
The employee who writes it down
An analyst notices that a safety check has been quietly skipped to meet a deadline. Telling her manager feels disloyal. Staying silent feels worse. Our questions are: what is the rule, who else can hear this without it becoming a career risk, what is the least damaging channel that actually works, and what protection does she have if it goes badly? The organisation’s test is different: if she had been wrong, would she be thanked for asking?
Where the method stops
This is not a device for justifying a decision that has already been taken. If the result is fixed and someone wants a document to dress it up, an ethical analysis becomes a costume. We ask openly whether that is the case. If it is, we say so, and offer to review the open question on its merits. If the answer is no and the pressure continues, we withdraw and keep the matter confidential.
We also decline or stop work that would require misleading people affected by a decision, crossing a legal or safety boundary, or disclosing confidences. A large benefit or a tight deadline does not move that line, for us or for anyone. Where a case is truly grey, we write the considerations down and put them to the person deciding, rather than settling them quietly ourselves.
We test ourselves against the opposite mistake: assuming bad faith, stereotyping, or reading emotions off a face. Inferences about what someone feels or wants are hypotheses and are marked as such. The client pays for the review, so we discuss any interests that could affect the judgement before taking on the work.
Discuss an ethical question
If your question involves a conflict of interest or a difficult ethical judgement, describe it without names or confidential details. You can also read about our approach and the Decision Review.