Agency PowerWhy Effective Control Matters More Than Formal Authority

Institutional power measures what someone controls in principle: the budget, the platform, the policy lever, the military. It is the upper bound of what could theoretically be done from a given position. Agency power measures something different — the fraction of that potential that can actually be deployed unilaterally, without being blocked, reversed, or absorbed by the institutions surrounding the actor.

This distinction is not theoretical. If you are trying to hold someone accountable — applying pressure through their incentives, building reputational cost, mobilizing regulatory attention — the relevant question is not how powerful they are in aggregate. It is how freely they can act. An actor with 95% agency and a power score of 80 is a more accessible accountability target than one with 50% agency and a power score of 95. The constrained actor can always point to their constraints. The unconstrained actor cannot.

Agency power answers a specific question: given what this person controls and how freely they can act, what is their actual capacity to change an outcome they have chosen to leave unchanged?

We apply distinct frameworks depending on whether the actor is a political leader or a corporate one, because the nature of constraint differs fundamentally between them.

For political leaders, the constraints are democratic and institutional: coalition requirements that demand consensus, independent judiciaries that can block executive action, legislatures that can override or defund, electoral cycles that impose regular accountability, and term limits that create finite windows of unchecked power. A leader governing through a fragile multi-party coalition — requiring active deal-making for every major initiative — operates at roughly half the effective power their institutional position implies. An absolute ruler with no functioning checks sits near 100%.

For corporate leaders, the constraints are structural and regulatory: voting control determines whether they can be removed; board independence determines how likely that removal is; regulatory exposure limits viable strategy; fiduciary obligations bind investment decisions regardless of personal conviction. A CEO with no controlling stake, facing active antitrust proceedings, and accountable to a genuinely independent board is meaningfully constrained — even if their company is worth two trillion dollars.

The standard model of corporate governance assumes that independent boards, activist shareholders, and regulatory oversight create meaningful checks on executive authority. In practice, these checks are often weaker than they appear — particularly for founder-CEOs who combine structural voting control with cultural authority that makes board challenge a reputational risk for the board itself.

Elon Musk owns X outright and controls SpaceX as a private company. His de facto authority over Tesla persists despite a sub-15% economic stake because the reconstituted board — assembled after his 2018 SEC consent decree expired — offers minimal independent resistance. Mark Zuckerberg controls approximately 60% of Meta's voting shares through a dual-class structure specifically designed to insulate him from shareholder accountability, despite holding only ~13% economic interest.

The regulatory layer is real but slow. Antitrust proceedings take years; compliance obligations cost money without necessarily changing strategic direction. The practical result is that many of the world's most influential corporate actors operate with agency factors above 0.80 — closer to the profile of consolidated political leaders than to a textbook public company CEO.

For high-agency actors — those above 0.80 — constraints are personal and reputational rather than institutional. The leverage points are incentive drivers: brand equity, capital market access, platform relationships, and the informal norms of elite networks. Pressure applied here reaches the decision-maker directly. There is no coalition to fracture, no court to appeal to, no electorate to mobilize separately.

For low-agency actors — those below 0.65 — the most efficient strategy is often not targeting the individual at all. Von der Leyen, Netanyahu, and Trump all face structural constraints that routinely absorb direct pressure: coalition partners who must be moved separately, courts that provide independent relief, oversight bodies with their own incentive structures. The accountability target in these cases is frequently the constraint mechanism itself.

Understanding this is not a minor refinement. It is the difference between effort that reaches the decision-maker and effort that dissipates. The accountability ecosystem consistently undersupplies two things: specificity about who is actually responsible, and clarity about whether that actor has the freedom to respond. Agency power is how we track the second.