Trust in the room doesn't contain a difficult CEO's effect. How the room handles disagreement does.

You are closing out a strategic decision with your executive team. Nobody objects. The meeting ends early, and you read the quiet as alignment. A new paper gives you a reason to ask what else that quiet could be.

The paper is Why good teams struggle under narcissistic CEOs?, published online in Management Decision in August 2026. Its authors are Bradley J. Olson of the Dhillon School of Business at the University of Lethbridge, Matteo Cristofaro of the University of Rome Tor Vergata, Yongjian Bao of the Dhillon School of Business (Calgary Campus), and Satyanarayana Parayitam of the University of Massachusetts Dartmouth. It is open access, and I read the abstract and roughly the first half of the full text. The supplementary materials and the rest of the article I did not reach.

The method matters for how far you lean on it. It is a cross-sectional survey of 115 senior members of executive teams in firms in Beijing, Shanghai and Nanjing, collected in 2022 from a convenience sample. Each respondent had taken part in a strategic decision with at least four executives, the CEO included. Each person is a single informant, and CEO narcissism was rated by other respondents, not by the CEO. It is a survey of perceptions. It is not an experiment, and it cannot show that one thing caused another.

The headline result is simple. CEO narcissism was positively associated with team conflict (b = 0.262) and the total effect on team effectiveness was negative (b = −0.206). Conflict carried part of that association, not all of it.

The more useful result is that the model has two stages and they do not behave alike. The authors put it this way in their abstract:

Upstream, and contrary to prediction, the narcissism–conflict association becomes stronger at higher levels of team trust and decision commitment, although both conditions are associated with lower overall conflict.

Abstract, Management Decision, as read on the publisher's page

Read that twice. Teams with more trust and more commitment to decisions had less conflict overall. But where the CEO was rated as more narcissistic, those same conditions went with a stronger link between narcissism and conflict. Downstream, the picture was different. Cooperative conflict management and agreement-seeking behavior weakened the link between conflict and effectiveness, which became non-significant at high levels of either one. So in this sample, trust did not shield the team from the CEO. How the team handled disagreement once it surfaced did the shielding.

Here is why this lands differently for the person at the top. In this study the CEO is the variable being measured, and the people doing the measuring are the ones who report to the CEO. You are the person least able to see this from where you sit. The authors note, in their section on the setting, that if norms in a high-power-distance context suppress reported conflict, their estimates would be conservative. My own addition: a calm room is not evidence of an agreeable room. Executives below you can read the cost of disagreeing with you far more precisely than you can read it.

The paper itself reads like a warning for the board as well. Its practical-implications table lists actions such as CEO 360-degree reviews and requiring executive-team sign-off on major strategic decisions. Those are the authors' recommendations, set out in a table, and I did not read the reasoning behind each one.

The part the research doesn't say, but that I'll say plainly: I would not read this as a verdict on any one chief executive. The measure is a set of adjectives rated by colleagues, not a clinical assessment, and the authors say so in their text. The finding about trust is the one I would hold most loosely. It ran against what the authors predicted, and they flag that the power for their mediation test is modest. Their sentence is that the indirect effect "should be read with corresponding caution." The data are from 2022, from one country, and from a convenience sample, and those cautions about generalizing are mine as much as theirs. What I take from it is narrower. If your team reports high trust, that is not the same as having a team that can disagree with you out loud.

One thing to do this week. Before the next major decision is called, ask each executive in the room to say where they disagree, or what would have to be true for them to disagree, and ask them to do it before you state your own position. Then watch what happens to the next ten minutes. That is a test, not a method, and it is my suggestion, not the study's.

How an executive team handles disagreement around the person at the top is the kind of question the Leadership Profile is for. You can start at the Wiring Map.