Rules
Part of Leadership ethics: how decisions reveal values
Leadership Ethics Trends: Three Questions to Test a Claim
Leadership ethics trends tested with three questions: does a claim narrow what a leader can say they did not know, add an audience, or change what is owed?
What a leader can get away with has narrowed, and what a leader can plausibly claim not to have known has narrowed faster. That is the whole of the trend. A claim about leadership ethics is worth acting on only if it passes a test.
Four shifts that are observable in your own organization over the last two or three years serve as the test cases. No figures are offered, because the point is the test, not a forecast.
What to take away
- A claimed trend earns attention if it narrows what you can claim not to know, adds an audience, or changes what you owe. Most claims do none of the three.
- Trends change exposure, visibility and scale. They almost never change the obligation, so a claim that you now owe less to someone should be treated with suspicion.
- A distributed team removes most of the signals you used to read conduct from, and most of the signals the team used to read you. Both have to be replaced deliberately.
- People increasingly leave over how a place is run, and say so publicly. The cost of a conduct failure now includes the people who watched it.
Three questions that sort a trend from conference material
You will be told about many more trends than any leader can act on. Three questions sort the ones that should change your behavior from the ones that are conference material.
- Does it change what you can plausibly claim not to know?Every genuine shift narrows that space. A trend that does not narrow it is probably not a trend in ethics; it may be a trend in something else.
- Does it change who is watching?New audiences (systems, witnesses, regulators, the next employer) change the cost of a decision without changing its content. If a claimed trend adds an audience, take it seriously.
- Does it change what you owe anyone?Almost never. Trends change the exposure, the visibility and the scale. They do not change the obligation.
The table applies the three questions to the four test cases below.
| Claimed shift | Narrows what you can claim not to know? | Adds an audience? | Changes what you owe? |
|---|---|---|---|
| Distributed teams hide conduct signals | Yes: you no longer see who went quiet | The team, now reading you through fewer signals | No |
| People leave over conduct and say so | Yes: the account exists whether or not you hear it | Yes: the next employer, public platforms, the board | No |
| The personal and professional line closes | Yes: outside conduct is read as evidence of how you lead | Yes: anyone who reads about it | No |
| The system's decision is yours | Yes: "the model did it" has no standing | Yes: systems and the people they treat | No |
Test case one: the team you cannot read
A team spread across time zones hides who went quiet in a meeting, who stayed late, who stopped talking to whom. The team also loses sight of how you reacted to bad news in the corridor.
The claim passes the first question and fails the third. Both directions of signal have to be rebuilt on purpose, because the accidental version is gone. Make routine one-to-one questions that surface conduct without interrogation the main channel, and ask them of everyone on a set schedule.
What you once did visibly, do audibly. When you decline an exception, say so where the team can see. When someone brought you bad news early, thank them where others can hear. The underlying mechanism, that people read the real rules from what they see happen, is in leadership ethics.
Test case two: departures over conduct
Staff once tolerated how a place was run, or left quietly. Now they leave and describe it: to the next employer, on public platforms, and in exit interviews the board reads.
A conduct failure that once cost a leader one relationship now costs the account of it, told by the person best placed to tell it.
The test says this is a change in the audience for your decisions, not in the decisions. The person you treated as expendable is now a witness with a platform, and in some cases a protected one. In the UK, a regulator's published whistleblowing policy notes that the Public Interest Disclosure Act 1998 generally protects workers and other stakeholders who disclose malpractice from suffering detriment.
Assume every departure will be described accurately by the person departing. Then run the checklist item that most leaders skip: who left in the last year, and would their honest reason embarrass you? If yes, fix the thing, not the leaver. The person weighing whether to go, and what to say, is the subject of employee ethics.
Test case three: the personal and professional line
What a leader does outside work is increasingly treated as evidence about how they lead. Whether this is fair is a separate argument. That it is happening is not.
The test marks this as an added audience, not a new obligation, so nothing dramatic is required. The standard is the one you would apply to anyone else in your organization, and the check is whether you would be comfortable with your team reading about it. Most leaders pass. The ones who do not already know which incident it is.
Test case four: the system's decision
The fourth case narrows what you can claim not to know and adds an audience, yet changes no duty. More of what an organization does to people is done by rules configured once and run at scale: who is flagged, who is prioritized, who is denied, who is charged what.
Owning an automated decision
- Name the person who owns the decision
- Ask what the rule does to the worst-treated
- Ask when someone last looked
- Ask counsel what law applies before review
The accountability has not moved. If your unit configured the rule, or bought the system, or accepted its output without asking, the decision is yours. For each automated decision your team owns, name the person who owns it as a decision. Ask what the rule does to the people it treats worst, and when someone last looked.
Where automated decisions affect individuals' access to services, employment or credit, the law in many places imposes specific obligations, and they differ. Ask counsel what applies before the review, not after a complaint.
What has not changed
The team reads the rules from what you do. The top performer's exception is the loudest decision you make. Bad news arrives in proportion to what happened to the last person who brought it.
Incentives beat policies. None of that has moved, and the shifts above mostly make each of them more visible, sooner, to more people.
The US Department of Justice's Evaluation of Corporate Compliance Programs asks how senior leaders, through their words and actions, have encouraged or discouraged compliance. That question was always the right one; the test cases above only make the answer easier to find.
The leader who does well over the next few years is not the one who has read the most about what is changing. It is the one whose conduct did not need the audience to be small.
Common questions
Should I stop using informal channels?
No. Use them as if they were formal, because they are. The habit of writing what you would be content to have read aloud is the whole adjustment, and it improves the writing.
How do I rebuild the signals with a distributed team?
Schedule what used to be accidental. Routine one-to-one questions on a calendar, decisions on exceptions stated where the team can see them, thanks for bad news given where others hear it.
Which of these shifts will reverse?
None, on any evidence available. Treat them as the conditions under which you will lead for the rest of your career, and calibrate accordingly.
