The Watershed Has Not Been Asked

What Washington decides about American water is being decided without us. A note from the Llano ditch.

I walked the Llano ditch this spring. The water was moving the way water moves in a ditch that has been moving water for a long time — slowly, with the small turbulences that say this is alive, under cottonwoods that have been drinking from it longer than the United States has had a constitution. The mayordomo who walked part of it with me did not need a clipboard. He knew where the headgate sticks, where the lateral takes off, where the bank slumps every spring, where to listen for the sound of water in the wrong place. He has known these things for forty years. His father knew them. His father's father knew them. None of this knowledge has ever been written down because none of it needs to be.

This is what an acequia is. Not a ditch. A four-hundred-year-old governance institution, with engineering older than the nation that contains it, operated by parciantes whose rights to its water predate American law by two centuries and federal law by three. The acequias of northern New Mexico are the oldest continuously functioning democratic water institutions in North America. They have outlasted the Spanish crown, the Mexican republic, the American territorial government, the railroads, the timber companies, the Forest Service, Los Alamos, and every developer who ever tried to dry them up. They are still moving water this morning, including the water in the Llano ditch behind the apartment where I am writing this.

I came to Rio Arriba County after two combat tours in countries where I watched the politics of water from the air. I came because the watershed I was born into is dying and this one is not. I came because four hundred years of acequia governance is the kind of thing a person should pay attention to while it still exists.

I came here, in part, to learn what it would mean to ask the watershed first.

In the next eighteen months, the United States will commit to siting roughly 1.7 trillion dollars in new data center infrastructure. The Federal Energy Regulatory Commission reports US data center capacity already at fifty gigawatts at the end of 2025, growing at twenty-four percent compounded annually. By 2030, Wood Mackenzie projects the figure to roughly double. Data centers will account for about sixty-eight percent of all electrical load growth in the United States through the rest of this decade. American utilities have announced one-point-four trillion dollars in capital expenditure to support this buildout — a twenty-seven percent capex surge over their prior baseline.

This is the largest discretionary infrastructure decision of the decade. It is being made without a map.

There is no federal siting framework. There is no national review of which watersheds can absorb the water. There is no review of which grids can support the power without new fossil generation. There is no review of which communities have the legal capacity to negotiate with hyperscale developers who arrive with non-disclosure agreements and three-letter law firms. Private capital picks a site. A county that lacks the technical staff to evaluate it signs an NDA. The aquifer is asked to negotiate after the concrete is poured.

This has produced, in the last two years, three visible outcomes. In Grant County, Washington, two decades of hydropower abundance ended in a temporary natural gas plant approved to keep data centers running through a cold Columbia River winter. In Loudoun County, Virginia, average residential electricity bills are projected to rise as much as thirty-seven dollars a month to underwrite infrastructure for facilities that employ fewer than a hundred and fifty people each. In the Colorado River basin — the most water-stressed major watershed in North America, in structural collapse — new hyperscale proposals continue to land on county desks as if the river were not in deficit.

None of these outcomes were inevitable. All were chosen by absence. When the federal government does not produce a siting framework, capital sites itself.

The acequia tradition would call this a failure of the repartimiento. The repartimiento is the ritual fair division of water — the meeting at the headgate where the parciantes agree on what gets diverted, in what order, with what consideration for who is downstream. The repartimiento is older than American water law. It is older than American law. It is, in some basins of northern New Mexico, older than the English language as spoken on this continent.

The premise of the repartimiento is simple: the watershed has standing. Before the engineer, before the developer, before the merchant, the water itself is consulted. Not anthropomorphically — through the people who have learned, across generations, what the watershed can and cannot give. The aquifer gets first vote, the community gets second, capital gets third. Inversion of how American infrastructure is currently sited. Same principle, four-century application.

If you tried to build a hyperscale data center on Llano land — and someone will try, eventually, because the dry power available in the Four Corners and the cool nights in the high desert make this country exactly the kind of country private capital will look at — the mayordomo would walk the lateral, look at the recharge data, look at the existing rights downstream, and call a meeting at the headgate. The parciantes would speak. The decision would be made by the people whose grandfathers' bones are in the cemetery up the road. The developer would either agree to the watershed's terms or leave.

That is the kind of consent the acequia has always practiced. It is also the kind of consent the federal government is structurally incapable of producing. The Department of Energy has no siting framework. The Federal Energy Regulatory Commission has no large-load transparency requirements. Congress has authorized no tax structure that distinguishes between hyperscale extraction and distributed compute paired with heat reuse. The watershed has not been asked.

This publication exists, in part, to ask it.

I do not claim the acequia tradition. I am learning it. I am a US Army veteran who has lived in this county for less than a decade and who will spend the rest of his life trying to be a good neighbor to people whose families have been here longer than the state. What I can do, from the position I actually occupy, is take what I am learning from the parciantes seriously enough to apply it to the largest infrastructure decision of my country's decade.

The water in the Llano ditch this morning has not been asked whether it can support a two-hundred-megawatt cooling load. The Rio Grande has not been asked whether it can absorb the evaporative losses of a hyperscale facility sited above Albuquerque. The Ogallala has not been asked whether it can survive another generation of industrial withdrawal. The Colorado has not been asked, because the Colorado is already saying no in the only language it has — by going dry. Washington is not listening, because Washington does not know how to listen to a watershed.

I am going to spend the next five weeks writing about what should be asked, who should ask it, what the watershed would say if it could speak in committee testimony, and what a national siting framework that took the watershed seriously would actually look like. The framework exists. The map exists. The math is not difficult. The thing that is missing is the political will to ask, and the cultural literacy to understand the answer when it comes.

The acequia families of Rio Arriba have been asking for four hundred years. Some of them are still alive. The repartimiento still meets. The water in the Llano is still moving. The institution that knows how to ask still exists.

It is time for the rest of the country to learn from it.