Who Holds the Leverage Between a Boss and a Subordinate?

Every piece of work a professional produces travels upward through one person before it reaches anyone who decides budgets, promotions or scope. The manager summarizes it for their own boss, judges which parts matter, and describes it in meetings the employee is absent from. The quality of the work is one variable. The route it takes is another, and the route belongs to a relationship with its own structure.

That structure has a clear intellectual starting point. In 1980, John Gabarro and John Kotter published "Managing Your Boss" in Harvard Business Review, and their central claim still organizes the subject. The relationship between a boss and a subordinate is a relationship of mutual dependence. Authority runs one way on the organization chart. Dependence runs both ways in practice.

Seen this way, managing upward becomes a structural competence: the skill of working effectively inside a flow of authority, information and trust that the subordinate inherits and can shape only in part. Three variables carry most of the explanatory weight, and each builds on the one before it.

Dependence as the source of leverage

The first variable is dependence itself. A subordinate needs the manager for resources, for protection when something goes wrong, and for access to the opportunities that shape a career. That half of the ledger is obvious to anyone who has held a job.

The other half is where the leverage sits. The manager needs the subordinate for execution, for accurate information about what is happening below, and for skills or capacity the manager personally lacks. A manager with ten direct reports and demanding targets from above can see only a fraction of the team's work and can do almost none of it. The manager's dependence is the subordinate's leverage.

That leverage has a particular shape. It grows when the subordinate makes the manager more effective at meeting the pressures the manager faces from their own boss, their peers and their metrics. Behavior that looks arbitrary from below, such as a sudden demand for weekly cost figures, usually makes sense once those pressures come into view. A subordinate who reads them accurately can present their work as the answer to problems the manager already has. Influence built this way is durable because it serves both parties and the organization at once.

Information as the channel

The second variable is information, and it follows directly from the first. Because a manager cannot watch most of the work, they experience it through a filtered stream: status updates, summaries, dashboards, a remark in a corridor, a slide in a quarterly review. The subordinate controls much of what enters that stream, when it arrives, and what context travels with it.

This makes the information filter the most consequential tool in the relationship. Used well, it gives the manager an accurate and early picture, enough to intervene while options remain open. Its central discipline is the principle that a manager should hear bad news first from the person responsible, and early enough to act. A delay reported three weeks ahead becomes a shared problem: the manager can rescope, add help, or reset expectations with their own boss. The same delay discovered through a colleague in a senior meeting becomes a question about every other report the subordinate has sent.

Framing belongs to the same discipline. Identical facts support different decisions depending on the baseline, the scope changes and the alternatives that surround them. A two-week slip against the original plan can also be an early delivery against a revised plan agreed after the scope grew. Honest framing supplies that context before the headline number arrives, so the decision-maker can interpret the facts accurately. The test of honesty is durability: a sound frame still holds after the manager learns every detail.

A third element completes the channel. Every issue a subordinate encounters falls into one of three bins: decisions the manager wants to make, developments the manager wants to know about, and matters the manager expects the subordinate to handle. Sorting well is escalation judgment. Too little escalation looks like concealment; too much suggests an inability to decide. The boundaries shift with the manager's pressures and with the subordinate's track record, which leads to the third variable.

Trust as the interpreter

Information arrives through a lens, and the lens is trust. The same update reads as reassuring from one person and as a warning sign from another. That difference comes from accumulated history.

Organizational research has a name for the terms that history rests on. Chris Argyris wrote about a psychological work contract in 1960, and Denise Rousseau's work from 1989 onward developed the psychological contract as each person's own beliefs about what the two sides owe each other. Between a boss and a subordinate, the unwritten terms usually include loyalty in public, discretion with sensitive matters, and keeping the manager informed before they face their own boss. A breach of those terms costs far more trust than the act alone seems to justify, because it changes the manager's forecast of what the subordinate will do next.

Trust also moves at two speeds. It rises slowly, through many small demonstrations of reliability, and it can fall sharply after one surprise. This asymmetry explains why the early report matters so much: each accurate, timely update adds a little to the balance, and a single discovered omission can draw down months of deposits.

The same lens clarifies difficult bosses. A micromanager usually signals a trust deficit, and the effective response reduces the manager's uncertainty through frequent, early, visible delivery. An absent manager usually signals a capacity problem, which calls for lateral relationships and independent sources of feedback. A manager who presents the team's output as their own is often responding to an organization that expects managers to represent their teams, where the line between representation and appropriation is genuinely ambiguous. Each of these is a structural problem with a personal face, and responses aimed at the cause work better than responses aimed at the personality.

The constant core across forms

Organizational form changes the tactics. In a matrix, a professional answers to two or more managers with competing metrics, and the useful move is to surface conflicting demands early and let the managers settle them together. In remote work, the ambient visibility of a shared office disappears, and deliberate visibility takes its place: structured updates, written decisions, scheduled conversations. In founder-led companies, the manager may be a founder of the kind Paul Graham described in his 2024 essay "Founder Mode", someone who stays closely involved deep inside the organization and moves priorities quickly. Across cultures, the norms for disagreement and bad news vary widely, and Erin Meyer's The Culture Map (2014) offers scales for locating a manager's expectations about feedback, decisions and hierarchy.

Each of these settings rearranges the surface. Underneath, the same three variables govern the outcome: dependence, information and trust. The matrix multiplies the dependencies. Remote work narrows the information channel. Founders compress the time in which trust gets tested. Cultural distance changes how information is read.

The practical consequence is a way of reading any working relationship with the person above. Map what each side needs from the other. Trace the route by which work becomes a picture in the manager's mind, and take responsibility for its accuracy and timing. Estimate where trust currently sits and what moves it. A professional who works with those three variables in view turns good work into recognized work, and does it in a way that makes the manager, the team and the organization better informed. That combination marks the boundary between upward management and politics: influence that improves the decisions made above, earned through the accuracy of what travels up.