Most organisations carry a problem they prefer not to name: the insecure leader—the senior figure who, under pressure, micromanages, withdraws, or bristles at feedback, and whose unresolved relational patterns quietly distort the work around them.
A recent Harvard Business Review article sets out a disciplined sequence for team members managing such a leader: regulate the leader’s state, relate through connection, then reason through the problem. The advice is sound, and the need for it is real. Insecurity at senior levels is common, and its costs—to coordination, to candour, to the people absorbing the volatility—are well documented. Learning to manage it is a genuine skill.
Yet every instruction runs in the same direction: from the member, toward the leader. The member regulates. The member relates. The member reasons. The work of steadying the relationship is assigned, in full, to the person with the least positional power in it—and it assumes the member is secure enough to do that regulating in the first place.
This is the half the conversation tends to miss. Managing up is one answer to insecure leadership. It is not the whole answer. It leaves two things unsaid: members need a secure base of their own, and the member best placed to manage up is rarely the one who most needs one.
This article is about the other half. Three threads structure it: what the managing-up advice gets right, and the assumption it rests on; why the same leadership lands differently across a team, and why members are owed a secure base, not handed the job of managing the leader; and how a team can hold a secure base between them when the leader cannot reliably provide it.
The Case for Managing Up
The case for managing up is strong, and it begins with an accurate reading of the problem.
The Harvard Business Review article—by Jeffrey Yip and Dritjon Gruda—makes two claims worth taking seriously: that insecure leadership is common, and that it is manageable. The figures are sobering. By one estimate, around a third of adults hold an insecure attachment orientation; a recent survey of senior executives found that most report symptoms of impostor syndrome.
Insecurity at the top is not rare. It is ordinary, and largely unspoken.
The contribution is to treat that insecurity as relational rather than fixed. A leader’s micromanagement, withdrawal, or defensiveness is read not as a character flaw to be endured but as a pattern of dysregulation that intensifies under pressure—and that a capable subordinate can work with.
The advice distinguishes two familiar forms, anxious and avoidant, and meets each with the same disciplined sequence: regulate the leader’s state, relate through connection, then reason through the problem. The reframing alone earns its place. It moves the member from labelling a boss as “difficult” to reading the state beneath the behaviour.
This answers a real question, and answers it well: how does a capable, steady person manage an insecure leader?
Yet the answer runs one way—from the member, toward the leader—and it begins with the member regulating: bringing steadiness to the relationship before they can lend it to anyone. That presumes a member steady enough to start. It says nothing about two prior questions: whether the member can, and whether the member should have to carry this at all.
Attachment Sits on Both Sides
The managing-up advice treats attachment as something the leader has. It is not.
Attachment is not a property of one party in a relationship—it operates on both sides of every working relationship at once, shaping what is provided and shaping what is received.
The recent evidence has gathered on the receiving side. A three-wave study of nearly 600 employees across more than a hundred teams found that followers higher in attachment security perceived their leaders as more transformational and thrived more at work; those higher in the insecure orientations—over-dependence or counter-dependence—perceived the same leadership less favourably and thrived less. Within the study, because members under the same leader still diverged, the difference lay with the follower, not the leader. The leadership was not the differentiating variable. Reception was.
A separate line of work sharpens the point. Wu and Parker found that identical secure-base support from a leader does not reach everyone the same way: for members higher in anxiety it works mainly by building a sense of competence; for those higher in avoidance, by preserving a sense of autonomy. Same provision, different route in.
These patterns describe a single pathway, which the author’s Attachment-Filtered Interplay model makes explicit: uniform leader signals pass through attachment-conditioned working models, where they are differentially weighted, before they register. Appraisal precedes perception. In the author’s research, anxious orientations weighted a leader’s availability most heavily, and avoidant orientations a leader’s receptiveness—a descriptive, preliminary pattern, but a consistent one.
If members filter the signals they receive, there is no principled reason leaders do not. London and Zobrist theorise that a leader’s own attachment shapes the signals they send; the symmetrical proposition—that it also shapes what they receive—is a reasonable extension the field has yet to test directly. An anxious leader may read a quiet member as rejection; an avoidant one may read a bid for reassurance as intrusion.
This bears directly on managing up. A member’s attempt to manage up is itself appraised—through the leader’s filter. And the member’s capacity to make that attempt is shaped by their own.
Two filters sit in every exchange, not one.
Members Deserve a Secure Base
Assigning the work of repair to the member carries a deeper problem. It casts the member as the manager of the leader’s state and forgets that the member is also someone who needs a secure base in order to do their own best work.
This is not the soft part of the argument. It is the better-evidenced one. The same research showing that secure-base support reaches members differently also shows it reaches the insecure ones most consequentially. Wu and Parker found that a leader’s availability, encouragement, and non-interference predicted whether employees acted proactively—and that this provision mattered more, not less, for members higher in anxiety and avoidance. For the anxious, it supplied the sense of competence they could not reliably supply themselves. For the avoidant, the autonomy that made engagement bearable.
The members least equipped to manufacture their own security are precisely the ones for whom receiving it changes the most.
Set that beside the managing-up frame and a difficulty appears. The advice asks members to regulate, relate, and reason with an insecure leader—work that depends on the member being steady enough to do it. The member best able to do that work is the secure one, who needs a secure base least. The member who most needs a secure base is the insecure one, least able to manufacture one from a leader who cannot reliably provide it. The obligation, as written, lands hardest where capacity is lowest.
So the question is not only whether members can manage up. It is why the burden sits with them at all.
A secure base is not a courtesy a leader extends downward when they happen to be secure. It is a condition members need—and, when it is missing, are owed.
A Secure Base the Team Can Hold
When a leader cannot reliably provide a secure base, the obvious correction is to demand that they become one. That correction fails.
Attachment is relatively stable—not readily changed in the short term, and least of all by intention. This is not a verdict on the leader; it is a fact about how relational orientations work. That leaves both one-directional remedies short. Members cannot reliably carry an insecure leader. Leaders cannot will their insecurity away.
The way through is to change the unit. The secure base need not rest on one person.
Attachment theory has long held that a group, not only an individual, can function as a source of security—something members return to under strain. Over time, a cohesive group can even strengthen the security of the individuals within it. What is far less settled is the sharper proposition this article needs: that a team can serve as a secure base in the moment, under pressure, in a way that shows up in the work. When researchers looked for that mechanism directly, it did not hold up cleanly. The team-as-secure-base is therefore offered here as a direction, not a finding.
Two things make it more than aspiration.
The first is that the team is already a relational resource. Wu and colleagues found that seeking feedback from peers is a workable strategy for insecurely attached members—though with a telling asymmetry: anxious members reach for it and benefit, while avoidant members are less likely to. Peers help. They help unevenly. That is why no single route to support suffices.
The second is the leader’s own awareness. A leader who can read their own filter can build around it. An avoidant leader who knows they read a bid for reassurance as intrusion can commit to structured reception, so a member’s approach is not left to chance. An anxious leader who knows they read quiet as rejection can build channels that make steadiness legible.
This is where the direction of obligation changes shape.
The team can help steady the leader—through named roles, predictable rhythms, depersonalised routes for raising and closing concerns—not as emotional labour demanded of the brave, but because a self-aware leader makes it safe to.
Managing up stops being a solo act of courage. It becomes a shared practice the leader enables.
A secure base, on this view, is not handed down by one person. It is held between several—and the leader’s contribution is to know their own part in how it forms.
What This Changes for Leaders
1. Read your own filter first.
A leader’s attachment shapes the signals they send and, in all likelihood, the ones they receive—which makes the leader’s own orientation one of the two filters in every exchange. The leader who cannot see how they tend to weight a quiet member, or one who reaches for reassurance, is missing half of what governs the relationship. Self-awareness is not a soft adjunct here. It is the precondition for everything that follows.
2. Make regulation a structure, not a feat of character.
If the team is to help steady an insecure dynamic, that work cannot depend on whichever member happens to be steadiest. It has to be built in—standing review points, more than one channel for surfacing risk, clear ownership of what gets raised—so that members who process relational risk differently each find a usable path, and the burden never collects on the most willing or the most senior. Peers help unevenly; the structure is what closes the gap.
3. Treat the insecure member as someone to resource, not someone to manage.
The managing-up frame asks, “How does this person handle me?” The more useful question runs the other way: “What does this person need from the relationship to contribute—and how do the team and I provide it?” The member who goes quiet, or reaches for reassurance, is not a management problem. They are telling a leader something about how the conditions are landing.
A Measured Claim
The argument leans on strands of uneven weight, and its boundaries are worth stating. The frameworks at its centre—the Available–Receptive–Responsive practices (ARR framework) and the Attachment-Filtered Interplay model—emerged from a single technology function under chronic time pressure, with a small, secure-leaning sample and a design built to specify a mechanism, not prove one. The patterns are descriptive associations, not causal laws.
The corroborating studies point the same way without standing in for one another: the follower-attachment work examined transformational leadership and thriving, the secure-base studies proactive behaviour, the peer-feedback work help-seeking—each consistent with the claim that attachment shapes how leadership is received, none measuring the safety pathway directly. Two further steps are extensions, not findings: that leaders filter what they receive as members do, and that a team can hold a secure base between them.
None of these dissolves the argument. They bound it.
Reception varies, it varies systematically, and a leader who cannot read it is working with half the picture.
Conclusion
The conversation about attachment and leadership has largely been an argument about which way the obligation runs. Should leaders be a secure base, or should members learn to manage up? Both answers are useful. Both run one way—and both keep attachment on a single side of a relationship that has two.
The evidence gathered points to a different starting place. Reception is not neutral. The same leadership reaches members differently, the difference follows relational lines, and it falls hardest on the members least able to manufacture security for themselves—the ones for whom receiving a secure base would matter most. Asking those members to carry an insecure leader inverts who can bear the weight.
When the person at the top cannot reliably be a secure base—and often they cannot simply decide to become one—the answer is not to load the task onto whoever is steadiest, or most junior. It is to build a base the team can hold between them: through structure, through plural routes to voice, and through a leader self-aware enough to understand their own part in how it forms.
The question shifts accordingly. Not “How do I manage my leader?” Nor “How do I become a secure one?” But “What would let this team hold a secure base between us—and what is my own orientation doing to help or prevent that?”
That question has no single answer. But for a team carrying an insecure leader, it may be the one worth asking first.
References
Bowlby, J. (1988) A secure base: clinical applications of attachment theory. London: Routledge.
David, T. and Shih, H.-A. (2025) ‘Securing success: Exploring attachment dynamics and psychological safety for adaptive behaviors in a military context’, Military psychology, 37(4), pp. 269–279.
Edmondson, A. (1999) ‘Psychological Safety and Learning Behavior in Work Teams’, Administrative science quarterly, 44(2), pp. 350–383.
Fraley, R.C. et al. (2011) ‘Patterns of Stability in Adult Attachment: An Empirical Test of Two Models of Continuity and Change’, Journal of personality and social psychology, 101(5), pp. 974–992.
Goel, R., Game, A.M. and Sanz-Vergel, A.I. (2026) ‘Are You Secure Enough to Follow? The Influence of Follower Attachment on Transformational Leadership Perceptions and Thriving at Work’, Journal of business and psychology, 41(2), pp. 285–302.
London, M. and Zobrist, C. (2025) ‘Leadership communication behaviors and attachment styles: theory for educating leaders to meet communication challenges’, Journal of leadership education, 24(2), pp. 156–171.
Mikulincer, M. and Shaver, P.R. (2017) ‘Augmenting the Sense of Attachment Security in Group Contexts: The Effects of a Responsive Leader and a Cohesive Group’, International journal of group psychotherapy, 67(2), pp. 161–175.
Nodjoumi, S. (2026) Enacting Psychological Safety Under Pressure: Secure-Base Leadership and Attachment-Filtered Appraisal in High-Tempo Teams. SSRN working paper. Available at: doi: 10.2139/ssrn.6516138.
Wu, C.-H. and Parker, S.K. (2017) ‘The role of leader support in facilitating proactive work behavior: a perspective from attachment theory’, Journal of Management, 43(4), pp. 1025–1049. doi: 10.1177/0149206314544745.
Wu, C.-H., Parker, S.K. and Jong, J.P. de (2014) ‘Feedback seeking from peers: A positive strategy for insecurely attached team-workers’, Human relations (New York), 67(4), pp. 441–464.
Yip, J. and Gruda, D. (2026) ‘How to manage an insecure leader’, Harvard Business Review, March–April, pp. 125–129.