Before dawn, the grid is already revising a promise.
Yesterday’s forecast became a schedule. Generators offered power. Utilities committed resources. Reserves were placed within reach.
Then the schedule met the world.
Homes wake earlier than expected. Factories start late. Wind arrives above forecast. Equipment warms unevenly. A generating unit can fail without consulting the market that priced its output.
Across the western electrical system, markets, automatic controls, balancing authorities, operators, contracts, and physical infrastructure keep adjusting around those changes. Some corrections happen in seconds. Others require a person to choose among available actions and issue an instruction.
There is no western emperor of electricity.
There is also no permission to leave the difference unresolved.
Electricity has no waiting room where an imbalance can sit until every affected institution agrees on a theory of fairness. The relation has to keep moving across machines, owners, standards, forecasts, contracts, weather, fuel, transmission, software, and people authorized to act.
If every component performs beautifully and the handoff fails, the lights do not grade partial credit on departmental competence.
Call this the coordination burden: the work required to turn separately situated actions into a continuable path instead of a collision, an abandoned handoff, or a pile of locally successful decisions that fail together.
Political arguments have a bad habit of making this work disappear.
We will abolish the hierarchy. Break up the bureaucracy. Free the market. Abolish the market. Decentralize everything. Nationalize everything. Give the users control. Give the experts control. Let communities handle it themselves.
Alright.
Who orders the replacement part?
Who remembers the obligation after the person who made it leaves?
Who schedules the work that has to happen before another piece can begin?
Who decides between two sensible plans that cannot both happen?
Who keeps enough reserve capacity around for the day when efficiency stops being the main problem?
Who deals with the person who takes the benefit of a shared arrangement and pushes the cost onto everybody else?
Those questions always survive the revolution.
Markets deserve their strongest defense.
Friedrich Hayek’s central point remains useful: important knowledge is dispersed. People know local conditions, shortages, opportunities, substitutes, and changing circumstances that no central office can completely collect in time. Prices can carry some of that information across strangers without requiring one planner to understand the whole system.
Management earns a defense for a different reason. Somebody has to preserve the schedule after the launch, train the replacement, order the part before it fails, settle the clash between two departments, and make sure tomorrow’s service does not depend on today’s hero.
Public administration earns its defense at the expensive address. A market can discover that serving one remote person costs more than serving fifty convenient ones. A public institution can decide that the remote address remains an address anyway. Records, routes, budgets, workers, facilities, and law can preserve a claim that would disappear if it had to win a fresh profitability contest every morning.
Hierarchy can compress decision time. Markets can coordinate dispersed knowledge. Administration can preserve continuity. Standards let strangers interoperate. Professions carry practice across organizations. Contracts, councils, cooperatives, protocols, mutual aid, and local institutions can coordinate without one comprehensive manager standing over them.
Civilization is crowded with useful coordination instruments.
It is also crowded with the problems those instruments produce.
That is the problem-problem.
A hospital can contain excellent clinicians, schedulers, pharmacists, insurers, social workers, and records systems and still make the patient carry the path among them.
Every institution may be doing its own job competently.
The patient becomes the little institution through which the big institutions communicate with one another.
The coordination burden did not disappear. It entered a sick body.
That is the test worth carrying into every proposal to tear an institution down. Removing an office does not remove the memory, maintenance, timing, scarcity, enforcement, or conflict it was carrying. If those functions receive no new carrier, somebody still inherits them.
Usually somebody with less money, less authority, less time, and less ability to refuse.
Freedom at the center can become abandonment at the edge.
The same standard cuts the other way.
An incumbent does not get to point at every useful function it performs and declare its own structure eternal.
Elinor Ostrom spent decades studying systems that complicate the fantasy that complex coordination requires one center. Shared resources can be governed through multiple decision centers, local rules, monitoring, sanctions, conflict resolution, cooperation, and institutions nested across scales.
That research does not give us a fairy tale about decentralization.
Someone still keeps the records.
Someone still watches the boundary.
Someone still deals with the defector.
Power can become quieter when it is distributed. It does not become unreal.
The question is therefore never simply whether coordination exists. Every institutional design coordinates somehow.
The question is who carries the work, who carries its cost, who gets to correct the carrier, and whether the function can survive replacing the institution currently performing it.
Artificial intelligence lowers the cost of coordination.
That is one of the most exciting things about it.
Institutions that currently fail to see one another may become capable of sharing context. Records can survive organizational boundaries. Small organizations can acquire analytical capacity once available only to large bureaucracies and firms. Supply, care, infrastructure, translation, scheduling, and public administration can become dramatically more responsive.
The same capability can make an incumbent vastly better at gathering the field beneath itself.
The coordinator works, so more institutions connect to it.
The shared record works, so more records enter it.
The interface works, so alternative interfaces disappear.
The planning system works, so more decisions begin there.
Eventually, the institution carrying the coordination burden can point to the burden itself as the reason nobody else may safely carry it.
That is where yesterday’s problem returns. The visible authoritarian says: I should rule because I am better.
The useful institution says something harder to dismiss:
You cannot safely replace me because everything depends on me.
Sometimes that statement will be true.
Our answer cannot be to smash the useful thing and hope coordination grows back.
It also cannot be to turn dependence into a permanent title.
The grid tightens again.
Generation drops while demand continues. Automatic controls move. Software recalculates. Earlier commitments become callable resources. An operator still has to select a present course.
Keeping every option open is not neutrality. Delay can become damaged equipment, interrupted service, a dark operating room, a stalled elevator, a home medical device losing power.
Someone has to close the move. So grant them enough authority to close it.
Then ask the question one level above the switch:
The coordination burden is real. Its present carrier is not sacred.
The danger begins when the closer also decides where its authority ends.