Most people in Seattle have never heard of the two treaties that govern the rivers their electricity and their flood safety depend on. One runs on the Skagit, the other on the Columbia. Together they move hundreds of millions of dollars a year across the Canadian border, and they were both negotiated in a world of 1960s and 1980s engineering assumptions that no longer hold. On one of them, Seattle is quietly winning. On the other, the United States has overpaid Canada for sixty years. And the city utility that operates the Skagit dams charges its own customers roughly three times what Tacoma charges for the same public hydropower.
This is an explainer and an accounting. The treaties are not secret, but the economics inside them almost never get laid out in plain numbers. When you do lay them out, a clear picture appears: the United States built its own dams, financed Canada's, pays Canada fairly for a real flood-control service, and on top of all of that keeps overpaying Canada for power on a formula both governments have already admitted is too generous.
The first thing to get straight is that these are two separate agreements on two separate rivers. They get bundled in strategy because they are the same US-Canada energy relationship, but on paper they are distinct.
| High Ross Treaty (Skagit) | Columbia River Treaty (Columbia) | |
|---|---|---|
| Signed | 1984 | 1964 |
| River | Skagit and Ross Lake, plus a Pend d'Oreille rider | Columbia |
| Dams | Ross, Diablo, Gorge (Seattle City Light) | Mica, Arrow (Keenleyside), Duncan, Libby, Grand Coulee |
| What is at stake | A below-market power deal, a $4 billion relicensing, the Seven Mile back-flood | Flood control for the lower Columbia, and a large power overpayment |
| US operator | Seattle City Light | Bonneville Power Administration and the US Army Corps of Engineers |
The High Ross Treaty settled a fight over whether Seattle would raise Ross Dam and flood the Skagit valley on the Canadian side. Seattle agreed never to raise it. In exchange, British Columbia delivers about 35.4 average megawatts of firm, winter-shaped hydropower to the Seattle load center, and Seattle pays for it.
Here is the part almost nobody has noticed. Seattle's capital payments under that deal ran from 1986 through 2020 only, and they are now finished. From 2021 to the treaty's end on January 1, 2066, Seattle pays only a small operations charge, on the order of $300,000 a year, for roughly 310,000 megawatt-hours of firm carbon-free hydropower that is worth many times that at today's market prices.
The High Ross deal today
Seattle pays about $0.3 million a year for power worth roughly $20 to $24 million a year at market. That is a 70 to 130 times annual return, and it runs to 2066.
The deal was not always this good. Through the 1980s and 1990s the fixed payment ran above market and Seattle was arguably overpaying Canada. The crossover came in the late 2000s. Now that the capital payments have ended, Seattle sits firmly on the winning side. The right move is to defend it. Every proposal to reopen and modernize the power payment would move Seattle's price up toward the open market. The status quo is the best outcome the city has.
There is a separate problem on the Skagit worth flagging, because it points the other way. The three Skagit dams are in the middle of a federal relicensing whose settlement runs to about $4 billion over the 50-year license. Roughly half of that is ordinary operations and dam capital the utility owes regardless. Inside the rest sits a fish-passage program capped near $979 million. That is close to a billion dollars of ratepayer money for trap-and-haul passage on three high mountain dams, in a river system where the science points at road-runoff chemistry, not dam passage, as the main killer of returning coho. The relicensing is still open for public comment, which is the moment to demand an independent cost-benefit review and a ratepayer cost cap before the license locks in around 2030.
The Columbia River Treaty is the bigger and older of the two. Canada built three large storage dams that regulate the river so that US dams downstream can generate more steadily. In exchange, Canada is paid half of that extra downstream power, the Canadian Entitlement, worth in the range of $229 to $335 million a year by the US government's own valuation.
The problem is how the payment is calculated. The formula credits Canadian storage as the first increment of storage added to the whole system, ranked ahead of every US dam built after 1961, including Libby and Dworshak. Storage has diminishing returns, so ranking Canada first hands it credit for the most valuable slice and inflates the payment. Recalculated honestly, ranking Canadian storage after the US dams that were actually built later, the real present-day benefit is roughly half.
This is not a fringe claim. The Congressional Research Service, a nonpartisan arm of the US Congress, states in report R43287 that removing the first-added benefit and a related surplus limit would make the entitlement "slightly less than half of current values." A modernization agreement reached in 2024 independently confirms it: the two governments agreed to cut the entitlement about 37 percent immediately and roughly 50 percent by 2033. By signing that, Canada conceded the old number was too high.
The Columbia overpayment
A last-added, fish-constrained recalculation cuts the entitlement roughly 45 to 55 percent, on the order of $115 to $170 million a year. Even after the 2024 deal, the corrected number likely still overpays by $50 to $100 million a year.
The Canadian grievance about the Columbia is real and emotional. Filling the four treaty reservoirs permanently flooded roughly 600 square kilometers of Canadian valley bottom, about three times the size of Kootenay Lake, and the Arrow Lakes reservoir alone displaced 23 communities and more than 2,000 people. British Columbia points to that sacrifice to justify what it is paid.
The financing tells a different story. When the treaty took effect in 1964, British Columbia did not pay for the dams out of its own pocket. It sold the first 30 years of the Canadian Entitlement to a consortium of 37 public and 4 private US utilities for a lump sum of $254.4 million, and used that American money to build Duncan, Arrow, and Mica. The United States separately paid Canada about $64 million for 60 years of assured flood control, and the US Army Corps built Libby Dam itself in Montana. In 1964 dollars that was roughly $318 million up front, on the order of $2.5 to $3 billion today, before a single dollar of ongoing entitlement was ever paid.
So American ratepayers financed the very dams that flooded those Canadian valleys, paid separately for the flood control, and built the one treaty dam on US soil. Canada was compensated at every step. The hardship the affected Canadian communities felt was real, but it was a failure of British Columbia to share the money with the people it displaced, which is why the province had to create the Columbia Basin Trust in 1995, thirty years late, to finally send benefits back to the region. That is Canada's internal accounting, not a debt the United States owes.
There is one place where Canada holds genuine leverage, and it is flood control. To understand it you have to understand that flood control is not about holding water. It is about keeping empty space ready. A reservoir can only catch a flood if it has room to catch it in, so real flood operation means drawing reservoirs down ahead of the spring snowmelt, then capturing the surge and releasing it slowly. That is the opposite of what a power operator wants, which is a full reservoir for maximum generation and the flexibility to sell when prices are high. The flood-control payment compensates Canada for running its reservoirs against its own power interest.
The stakes are not hypothetical. The entire arrangement exists because of one disaster. On May 30, 1948, the Columbia, running at twice its normal volume, breached a levee and destroyed Vanport, Oregon, then the state's second-largest city of about 18,500 people, in a single afternoon. At least 15 people died. That flood is why the United States wanted Canadian storage in the first place. The risk concentrates downstream, at the mouth, where the whole basin's flow converges on the flat, densely built floodplain of Portland and Vancouver. Today Canada holds back more than a trillion gallons each spring for US flood protection.
As of 2024 that flood service shifted from an assured, pre-paid arrangement to a "called upon" one, where the United States requests Canadian flood storage and pays for it. The interim version of that arrangement expires July 31, 2027 without a ratified treaty. That deadline is the pressure point, and it is the reason the ports of Portland and Vancouver have been sounding the alarm.
The clearest way to see the whole thing is to keep two ledgers separate, which is exactly what Canada prefers to blur.
The first ledger is power. The United States overpays because the entitlement is calculated on an inflated 1960s formula. That overpayment is on the order of $50 to $170 million a year, it is indefensible, and both a US government report and Canada's own 2024 signature say so. This is the ledger to correct.
The second ledger is flood control. Canada genuinely delivers it, at real cost to its own power business, on land it flooded to build the reservoirs. It is worth paying for, at about $37.6 million a year plus call costs. This is the ledger to honor.
The thing the United States should pay for is real and modest. The thing it should stop paying is inflated and several times larger. Splitting the two is the entire game. It is also clean: the United States is not refusing to pay Canada for flood control, it is ending a six-decade overcharge on power that a US report and Canada's own signature both call too high. Money that stays home instead of flowing north.
All of this is federal, run through the State Department, the Bonneville Power Administration, and the Army Corps, not through City Hall. But there is a local hook, and it is a sharp one. Seattle City Light operates the Skagit dams and sits on some of the cheapest power on the continent, including the below-market High Ross deal. Yet it charges its residential customers about 13.4 cents per kilowatt-hour, roughly three times Tacoma Power's 4.5 cents, for the same publicly owned Northwest hydropower, while raising rates again.
The treaty economics and the utility economics point at the same conclusion. Seattle has access to extraordinarily cheap public power, and both the federal government and its own city utility have arranged for ratepayers to see far too little of the benefit. The federal fix is to stop overpaying Canada for power while honoring the flood-control bill. The local fix is for Seattle City Light to justify a threefold markup on some of the cheapest power in the country, or pass it through.
None of this is aggression toward Canada. It is fairness. The United States built its own dams, financed Canada's, and pays a real bill for flood control. What it should stop doing is overpaying, forever, for power on a formula that everyone involved already agrees was too generous.
Want to support this operation, contribute expertise, or join as a member? Tell us who you are and how you want to help.