The most ethical leaders aren't the ones who deny power imbalance exits. They're the ones who plan around it — for the people navigating it from the other side.
Dr. Alvarez liked to think of his lab as flat.
Open door. Informal culture. First names. Encouraged debate — he said this explicitly, more than once, to new trainees and to the lab as a whole. He meant it. He genuinely wanted to know when people disagreed with him, when something wasn't working, when a direction he had chosen was wrong. He thought he had built a lab where those things could be said.
Then, during a feedback session, a trainee said something that unsettled him.
"I didn't feel like I could say no."
Dr. Alvarez sat with that for a moment before responding. He hadn't pressured anyone. He hadn't raised his voice, hadn't expressed disappointment, hadn't done anything that would appear on a list of things not to do. The ask had been reasonable. His manner had been collegial. He had genuinely believed the trainee was fine with it.
But the trainee hadn't been fine with it. And they hadn't said so.
What He Had Misunderstood About Power
Dr. Alvarez's assumption — the one that the trainee's words had quietly dismantled — was that power needed to be actively wielded to be felt. That in the absence of coercion, pressure, or explicit expectation, the hierarchy between a PI and a trainee effectively collapsed. That an open door and an informal culture were sufficient to create genuine psychological safety.
They weren't. And the reason they weren't had nothing to do with his intentions.
Power in a research lab doesn't come only from behavior. It comes from structure. Dr. Alvarez wrote the letters of recommendation that would shape his trainees' careers. He decided who was included in projects and who wasn't. He controlled access to resources, to authorship, to the opportunities that determined what options his trainees would have when they left. None of this required him to be threatening or demanding or even particularly directive. It simply existed — as the structural reality of the relationship — and it shaped what his trainees felt they could safely say, regardless of what he had told them about the open door.
The trainee who said "I didn't feel like I could say no" was not describing a failure of his intentions. They were describing the gap between intentions and structure — a gap that good intentions alone cannot close.
What He Started Noticing
Once Dr. Alvarez understood what he had been missing, he couldn't stop seeing it.
He started paying attention to patterns he had previously read as personality differences. Who spoke freely in meetings and who hedged. Who volunteered dissent and who waited until after the meeting to say quietly what they had not said publicly. Who absorbed extra work without complaint — not because the work was fine, but because complaining felt riskier than absorbing. Who said yes quickly to things that, on reflection, were not actually their responsibility to say yes to.
These patterns were not random. They tracked, fairly consistently, with who had the most to lose. The trainees earliest in their careers, most dependent on his recommendation, least certain of their standing — these were the ones hedging, waiting, absorbing, saying yes. The ones with more security — a fellowship, an offer elsewhere, a publication record that gave them options — were the ones who pushed back, asked hard questions, said no when no was the right answer.
The open door had not been equally open to everyone. It had been most open to the people who needed it least.
The Habit: Acknowledge The Power Imbalance. Make It Safe To Disagree.
Dr. Alvarez couldn't eliminate the power differential. That was structural, and it wasn't going away. What he could do was acknowledge it explicitly and build around it — reduce the hidden risk that his trainees were navigating without any map.
He started making consequences explicit in situations where his trainees might reasonably be uncertain about them.
"Saying no to this won't affect your evaluation or your standing with me. I want to know if it's too much."
"I'm going to ask for volunteers on this. If nobody wants it, that's genuinely fine — I'll find another approach."
"I'm raising this as a question, not a direction. I actually don't know the right answer here and I want your honest read."
These statements were not reassurances meant to make people feel better. They were specific pieces of information designed to address specific uncertainties that the power dynamic created. A trainee who doesn't know whether saying no will cost them something will often say yes rather than find out. A trainee who has been told directly that no is safe has a different calculation available.
He also changed how he collected feedback and dissent. A lab meeting is not an equal environment for all voices — it is a room where the person with the most power sits at the head of the table and everyone else calculates, in real time, how much disagreement is safe to express in front of their peers and their PI simultaneously. Dr. Alvarez started creating alternatives. One-on-ones where concerns could be raised privately. Anonymous written feedback for certain kinds of questions. Direct, specific invitations for individual trainees whose perspectives he noticed were going unheard in group settings.
He also stopped reading silence as agreement. This was perhaps the most significant change — because silence, in a power-imbalanced room, is almost never neutral. It is the sound of someone calculating risk. Treating it as consent was not just inaccurate. It was a way of using the power differential to arrive at outcomes nobody had actually agreed to.
What Shifted
The change wasn't dramatic or immediate. Power differentials don't dissolve because a PI says the right things — they are structural, and they persist.
But something shifted in what Dr. Alvarez was able to know about his own lab.
People started telling him things they hadn't told him before. Not because the power had gone away, but because the cost of honesty had been reduced enough to make honesty feel viable. A trainee who was overloaded said so. A postdoc who disagreed with a methodological direction said so, in a meeting, without waiting to see if it was safe. A student who needed to say no to a request said no — clearly, early, and without apparent anxiety about what it would cost them.
Dr. Alvarez had not become a different kind of person. He had built a different kind of environment — one in which the power he held was acknowledged rather than denied, and the risks it created for the people below him in the hierarchy were actively mitigated rather than left for them to manage alone.
The most ethical thing he could do with the power he had, he came to understand, was not to pretend it wasn't there. It was to plan around it — for the people who had no choice but to navigate it.
What This Habit Asks of You
Three practices, any of which can begin immediately.
Make the stakes explicit when they're uncertain. When you ask for a volunteer, when you make a request, when you invite feedback on something you care about — say clearly what the consequences of different responses actually are. Not as a disclaimer but as genuine information. Your trainees are doing a risk calculation you may not be aware of. Give them accurate inputs.
Create multiple channels for honesty. Not everyone can say the difficult thing in a room full of people, in front of you, in real time. A lab culture that only supports public, synchronous dissent will hear from the people with the least to lose. Build in private options — one-on-ones, written feedback, direct individual invitations to share a perspective. The goal is to hear from the people who have something real to say, not just the ones who feel safe saying it loudly.
Stop equating silence with agreement. When a room goes quiet after a proposal, when nobody volunteers concerns, when feedback is uniformly positive — treat this as data requiring interpretation, not as confirmation. Ask specifically. Ask privately. Ask in formats that reduce the risk of saying the thing that the room wasn't saying. You will often find that the silence was not agreement.
A Note on What This Isn't
This is not a habit about eliminating authority or pretending that the PI-trainee relationship is something other than what it is. Dr. Alvarez still had power. He still made evaluative judgments. He still controlled access to opportunities. None of that changed.
What changed was his relationship to that power — from ignoring it to accounting for it. From assuming that good intentions neutralized structural hierarchy to understanding that structural hierarchy persists regardless of intentions and requires active mitigation.
An open door is not the same as a safe room. An informal culture is not the same as a flat one. The most ethical leaders in research are not the ones who believe they have no power. They are the ones who take that power seriously enough to plan around it — for the people who are navigating it from the other side.
You don't have to misuse power for it to shape everything in the room. It shapes everything by existing.
The question is not whether your trainees are aware of the hierarchy. They are — far more acutely than you may realize, because they have more reason to track it carefully. The question is whether you are as aware of it as they are, and whether you are using that awareness to make the environment more honest rather than less.
That's not a hack. That's a habit.
✨ The GradLab Compass Trainee-Mentor Meeting & Feedback System include tools for raising concerns upward, framing pushback as shared goals, and saying no clearly and early — built for the actual dynamics of research environments.
Stories are fictionalized or composite narratives, created to illustrate common challenges and patterns in research life. They are intended for educational and reflective purposes and do not represent any specific individual or institution. 