Organizations often say they want people to have more agency. Usually they mean initiative: notice the problem, make a recommendation, move without waiting to be told.

But agency and authority are different things. Agency is a person’s capacity to perceive, choose, and act. Authority is the system’s permission for that action to carry consequence.

A person can have enormous agency and very little authority. They can see the pattern, understand the tradeoff, and know what should change while still needing approval from someone with less context.

This is why exhortations to ‘be empowered’ can become frustrating. If decision rights, information access, budgets, incentives, or consequences have not changed, the architecture has not changed either.

The person may be invited to act while still being punished for crossing a boundary the organization never made explicit. Or they may be asked to own an outcome without control over the conditions that produce it.

Repeated escalation is a clue

When the same category of decision repeatedly moves upward, I become curious. Sometimes escalation is appropriate. Some decisions carry enterprise risk and should be made centrally.

But repeated escalation can also mean authority lives too far from the information required to use it well. Senior leaders become routing layers, while the people closest to the work learn to wait.

The goal is not maximum autonomy. It is alignment between agency, information, authority, and consequence.

Distributed authority requires trust, but trust is not simply a feeling leaders either have or do not have. Systems can make trustworthy action easier.

Clear principles, visible constraints, accessible information, feedback loops, and explicit decision rights allow people to act without guessing what will later be considered acceptable.

If you want more agency, do not only ask people to behave differently. Look at what the organization allows them to know, decide, spend, change, and learn from.

People can see what needs to change long before they have permission to change it. The architecture determines whether that seeing becomes useful.

When people do not understand the process, the expectations, or the rules, they cannot meaningfully influence the outcome. Ambiguity can be useful in exploration, but hidden rules are not the same thing as freedom.

I am interested in systems where people can see enough of the architecture to locate themselves inside it: what is fixed, what is negotiable, what they own, what someone else owns, and where a decision can actually be changed.

Agency requires more than permission to act. It requires enough legibility to act intentionally.