A town was told to prove it had room. The proof is the point.

Washington's Department of Commerce appealed Duvall's twenty-year growth plan in 2025, saying the city had not shown room for enough housing, especially for people with lower incomes. That sounds like a fight about whether to build. It is not. Duvall has added 514 homes since 2019, about 73 a year against the 36 a year its target implies. The city's own count of room left in the zoning is 932 homes against a target of 890. The whole fight is about price: room for 675 homes for households under 80 percent of the county median, 409 of them under 30 percent, which for a household of three is about $44,400 a year.

The city's consultant showed that room with a rule: any zone that allows twenty or more homes an acre counts as housing for households from zero to eighty percent of median, the band that holds all 409 of the deepest-need homes. No subsidy enters that arithmetic. No operator, no covenant, no funder, no rent. Before the plan was adopted, the city's own senior planner had written to the state that "it would not be genuine to equate high-density housing with affordable housing." The analysis then does exactly that.

You do not have to take that characterization from us. Here is city staff explaining the exercise to the planning commission on August 26, from the meeting's own recording:

That is an honest description of what the law asks, delivered by people doing their jobs. The seats are the homes. A new two-bedroom apartment in Duvall needs about $2,300 a month to be worth building, by our estimate; a household at 30 percent of median can pay about $1,109. Zoning cannot close that gap. Public money can, and public land can, and otherwise nothing does. Duvall's entire general fund for the next two years is about $16 million, and it pays for the police, the parks, the streets and the staff. The metric will be met. Not one of those homes is funded. Both sentences are true, and only the first one goes in the file.

If that were the only instance it would be a quirk of one statute. It is not. Read the rest of the town's paperwork and the same move turns up twice more, each time an obligation met in form while its purpose waits.

In the word "affordable." The city's draft affordable-housing chapter calls a home affordable if it rents for what a household at 80 percent of median can pay: about $2,425 a month for a two-bedroom, before utilities. The valley's newest apartments, built in Carnation in 2021, list two-bedrooms from $2,196, and the typical starting rent across 39 nearby buildings is about $2,300. A home "restricted" at 80 percent could charge more than the market already does. The restriction would bind on nothing. State law saw this coming: for a program like the city's, the default ceiling for rentals is 50 percent of median, and a city may go to 80 only after a public hearing and a written finding that local conditions need it. Neither has happened. At 50 percent the same restriction cuts about $546 a month from each restricted home. The state's own auditors found the pattern statewide in 2019: the legal low-income ceiling sat above typical market rent in at least 498 zip codes.

In the public's own land. Duvall's plan has carried a policy for years to "identify and promote the use of surplus public-owned land for housing affordable to low- and moderate-income households." In 2023 the city reviewed that policy and wrote down the result: "A few years back, the city rezoned surplus land from R4 to Public due to outcry from adjacent neighbors." The one action taken under it moved public land away from housing. The same 2023 plan proposed a ground-lease program again, with an inventory of buildable public land marked short term. The September 2026 draft carries it forward as something to "analyze" and "explore." Meanwhile the city holds a vacant two-acre lot on Big Rock Road, handed over by a developer in 2022 under a 2008 agreement "to establish a city facility," appraised by the county at about $3 million, and sitting in a proposed zone where the city's draft code requires homes in any new building.

A seat that fits and will never be built. A ceiling set above the market. A policy promised three times and performed once, backwards. Nobody lied in any of them, and each one can be filed as done. That is the pattern this page is about: a metric standing in for an outcome. The rest of the page is about how it spreads, up a ladder of institutions none of which answers for the whole, and ahead of the rules themselves.

One river, one valley floor, and effectively two ways out of town. Here is a resident at the August 26 meeting, describing what the plan is being written on top of:

Ninety-two percent of Duvall's working residents work outside the city. Last December the Snoqualmie reached Flood Phase 4, the county's highest, the state highway closed north of town, a fallen tree closed Woodinville-Duvall Road, and on December 11 Duvall and Carnation were cut off. By standard rates the rezone showing implies about fifteen hundred more parking spaces in a town with no garage. The capacity analysis carries no road assumption and no flood assumption. On September 15 staff said the traffic model is being updated to the new capacity, and that "once that's done, we're going to look at levels of service." That result is not published. The arithmetic is not about this place. It is about whether the seats fit.

Count the links between the statute and the family it names. The legislature writes the law. A state agency writes the guidance that decides what counts as capacity for an income band. The county allocates targets to each city by income band. A regional council certifies the transportation half. The city writes code to match. One consultant picks the model parameters that decide whether the showing passes, and another drafts the code language. A developer, the only actor in the chain who can build anything, builds where it pencils.

Every link is legitimate, and most are accountable to somebody. The problem is the sum. No resident of this town can vote out the state agency. The county council answers to a county in which the town is a rounding error. The regional council is a body of governments rather than of voters. The consultants answer to a contract. What a resident gets is a comment period on a code amendment whose parameters were fixed four links upstream, in rooms that were open, minuted, and effectively out of reach.

Follow one rung and you can watch it work. The state would not accept the city's work plan until King County signed off on the new capacity analysis, and the county did, last December. On the council's January 6 recording, staff report that "King County identified" that the new analysis "doesn't require us to annex any land ... We can do it within the city limits." So the answer became rezoning inside the existing town. The city's own adopted plan points somewhere else as well: a new growth area south of Big Rock Road, "to meet growth targets and accommodate deeply affordable housing," to be requested "when the County is ready to process requests." That door is the county's to open. The same season, county conservation money helped the city buy three acres at the edge of town that "will remain undeveloped." Each of these is lawful and defensible. Together they decide where the town's affordable homes can go, and nobody put them on one page.

This is not a conspiracy, and calling it one is how people lose the argument. It is a slope: each step defensible on its own terms, the whole accountable to no one in particular. That is why it is durable, and why anger slides off it.

The pressure is real and documented. The state asked the hearings board to declare the plan invalid. On a finding of noncompliance the governor may withhold a city's share of fuel tax, sales tax and liquor revenue, and suspend its authority to collect the real estate excise tax. A small city cannot absorb that, so it complies, on a deadline, against a model it cannot test.

Then comes the part that is easy to miss, and here the city says it in its own words. On January 6 a council member asked whether the upzones in the town's lowest-density neighborhoods were needed at all. Staff answered that apart from how zoning blends between areas, "we do not need to change ... anything lower than our R20 really." Later the same evening staff explained why they were in the plan anyway. They were not expected to "yield an affordable unit, but they would satisfy a niche, which is another box that we can check with Commerce. That is the missing middle. We're not required to do those housing bill requirements ... because of our population size." They were there because the city expects to be asked later, and "if we kind of do the work early ... it makes it much easier for us to deal with things in the future." On September 15 staff put it plainly: middle housing "is not required to meet capacity."

The planning press calls this the shadow effect: the fallback exists, so the local decision bends before anyone uses it. We do not claim the state made Duvall do this; the city's own account is that it chose to get ahead of a rule it expects. That is the shape exactly. Eighteen states now limit local authority over accessory dwellings alone. Oregon preempted local zoning for middle housing in 2019, Florida and California have their own versions, and in each case behavior changes ahead of the law. To its credit, Duvall's council has started to push back: on September 15 it leaned toward a narrower local definition of middle housing, and a worksheet is coming.

Hold that sentence. It is how the pattern spreads, and the next section is the same shape wearing different clothes.

Start where it touches ground rather than where the pundits are. Two thirds of planned data centers are going to rural areas, and nearly four in ten to counties that have none today. Across the country, county commissions and city councils have been doing what councils do: a county west of Houston paused new AI data centers in July, a Nebraska county suspended permits in May, a Missouri city passed a one-year moratorium in August, and one industry tracker counted seventy-five projects worth about a hundred and thirty billion dollars disrupted by local opposition in a single quarter. The number most quoted in those hearings is a rate spike in one regional power market, because the thing people show up for is the electricity bill. The AI industry is not an abstraction arriving later. It is a land-use actor now, in counties that look a great deal like this one. If a town ever puts such a room in a building of its own, it should be a room and not a campus, named for what it does, its heat put into the building's hot water, and the first thing switched off when the power fails.

And the answer arrives in the familiar shape. Georgia's HB 1012 would bar counties and cities from issuing data center permits at all for a period, and this month state legislatures began moving toward postures that had until now been purely local. Whichever way a given bill points, the decision is moving up and away from the county commission. That is the ladder on this page again, being built in a year rather than over thirty.

The count can be argued, and parts of it can be won. The city publishes its room by zone, not by income, though income is what the appeal turns on. Using the consultant's own method, we estimate room for 665 to 719 homes under 80 percent of median against a need of 675: anywhere from 10 short to 44 over. So "we need everything that is left" is a fair answer about the dense zones. It is not an answer about the neighborhoods, where the plan adds 177 homes of room for an income band that needs 10. Publishing the city's own table settles it in an afternoon.

But win the count outright and nobody is housed, because the count moves no money. The question underneath it is about money and land, and it can be answered. The two cheapest tools a small town holds need no upzoning at all. The first is waiving its own fees on homes that are really restricted: city fees run about $33,000 an apartment, a September 8 memo from the city administrator sets out how the council could waive most of them, and at a 50 percent ceiling the waiver buys about $1.57 of rent reduction for every public dollar, against about $0.81 for the tax exemption the city is drafting. The second is putting public land under the homes on a ground lease.

The lot on Big Rock Road is the worked case. At its own appraisal the land is worth about $48,700 for each apartment it could carry. Waived fees and the tax exemption add about $97,000. Together that is about $146,000, and holding a rent at 30 percent of median for twelve years costs about $123,000. Fees and the exemption alone fall short. With the land, they do not. Nothing else on the city's list can do that.

The city's consultants have proposed a new police station on that lot, or a police station and city hall together. No decision has been made. In our view the lot should not become a police station, and the case for a new station is weaker than it sounds: the city renovated the one it has in 2023 and 2024, the problems named a year later read as remodel work, and the consultants themselves warned that a new station and city hall would cost more than the council can borrow without asking the voters. The city's own plan already describes a better building for the site: shared workshops, a kitchen and teaching space at street level to support family-wage jobs and new businesses, permanently affordable homes above on a city ground lease, and a solar roof with backup power so the south end has somewhere to gather when the roads close.

The same question stands one level up. Every actor on that ladder spends public money and every one of them is short. The county that assigned this town its target bought the valley's farming future in 1986 and a 75-acre farm on Duvall-Monroe Road in 2020 for just over a million dollars, for a trail extension ranked thirty-first and unfunded at somewhere between seventeen and thirty-five million, and it runs the lease program that would put that ground back to work on about one and a fifth staff positions. That farm lies outside the city limits and adds nothing to Duvall's housing numbers; we say so first. A county that asks a small town to make room for homes it cannot fund can be asked to put its own idle ground to work. The state's instrument of compulsion is not a fine; it is withholding the town's share of fuel tax, sales tax and liquor revenue. The entire chain runs on fiscal power, and not one link in it is flush. The same is true one level down: a city that keeps an unpublished list of surplus property and a promise three times deferred can be asked to weigh its own lot in the open.

The numbers at the top of that chain are public and they are not small. Federal agencies reported about $186 billion in improper payments in fiscal 2025 alone, across sixty-four programs, an increase of twenty-four billion on the year before. The Government Accountability Office puts total annual losses to fraud somewhere between $233 billion and $521 billion, a range that wide because nobody actually knows. Those are the auditor's own figures rather than an activist's, and the same office has published on using machine analysis to find them, for the plain reason that the volume of transactions passed what human review can cover a long time ago.

What has changed is that checking became possible again. Reading a budget, a contract file, a capacity memo or a payment register the way an auditor reads it no longer takes a department, and that is the whole of the opportunity here: the ledger can be made to run the other way, with the institution legible to the people it bills rather than the people made legible to the institution. It is not a radical proposition and it does not belong to either side. One side has been saying for a generation that government wastes money, and the other has been saying that public goods deserve honest stewardship. Both claims are checkable now, by the same method, and neither has any reason to fear the answer.

That is a better use of a comment period than one more argument about density. Ask the city and the county what they spent, on what, and with what result, what land they hold and what it is for, and ask for it in a form somebody outside the building can actually check. It is the only demand on this page that gets stronger the more money is involved.

In February 2025 a nonprofit and two Mercer Island residents appealed that city's plan to the state's Growth Management Hearings Board. That August the Board agreed with them and ordered the city to show how many homes would actually be produced at each income level, not only how many could fit, and to adopt protections against displacement. Mercer Island complied rather than fight. In July 2026 its council adopted affordable set-asides at 50 percent of median for rentals, a fee in lieu that funds homes below 50 percent, public sites identified for housing, and notice and relocation help for tenants priced out. The Board will look past paper capacity and ask what gets produced, and the package a city ends up adopting looks a lot like the one on this page.

It would be easy to read all of the above as an argument about paperwork. Here is the month it is happening in. The grocery index for the year to August came in at 2.7 percent, which sounds calm and is being reported that way. It sounds calm because the move already happened: food at home is up about thirty-two percent since January 2020, roughly twelve to fourteen hundred dollars a year on an average household's bill. A low rate sitting on a floor that already jumped is not relief. It is the new floor being described as stability.

And the next push is in the pipe rather than in the index. When the Strait closed in the spring, Gulf producers did not merely raise prices, they declared force majeure and suspended export contracts outright, and the transmission from fertilizer to food runs about six to nine months, which puts the effect in late 2026 and early 2027. Underneath all of it, diesel set an all-time national record on September 14 at $6.23 a gallon, up about sixty percent since late February, and it stayed high while crude retraced because the constraint moved from the barrel to the refinery. Diesel is the one input that touches every item on the shelf, because food is grown with it and then moved with it. The deeper page follows that chain properly.

Nobody is lying and none of those numbers is false. The measurement simply sits downstream of the event, and right now the interval is long and the pipeline is full. A household that looks at that and decides to stock up is not panicking; it is reading a lag correctly, which is the only move available to people who cannot hedge. That is the same move the capacity showing makes, and the same move we are asking a council to stop making.

It also settles what the public's land should be doing. At the end of a road that closes, the thing worth having is the capacity to grow food here, to make things here, and to house the people who do both. The public already owns seventy-five acres of the first, outside town, and two acres of the other two, inside it. That is not nostalgia. With diesel at a record and the fertilizer bill arriving with the harvest, the distance your food and your job travel is a line item, and it is the line item a comment period could actually move.

In Washington a zoning ordinance cannot be put to a local vote; the courts hold the zoning power belongs to the council. Petitions persuade, they do not decide. These do more:

1. Ask for the numbers as a legal demand. A public records request for the capacity workbook by income level, the displacement analysis, the list of surplus city property, and the site study for the new city buildings. The law starts a five-business-day clock.
2. Put a comment in the record, by email to the city clerk, by noon on the day of the meeting.
3. Comment on the environmental review of the plan and code amendments, due October 2.
4. Propose an amendment, not only an objection: in the plan and code under review until adoption, with the budget in October and November, or on the annual docket.
5. Speak to the money at the budget meetings on October 6 and 20 and the final hearing on November 3.
6. Write to the state reviewer during its sixty-day review, where compliance is actually judged.
7. Take part in writing or in person, because only someone who did can appeal to the Growth Management Hearings Board, within sixty days of the ordinance.
8. Use the county's doors: the growth planning council meets November 18.

The campaign that grew out of this work sets out every one of these with its date, its address and its statute, and publishes one letter that does five of them at once: dontbothellduvall.com. Its letter, its record of the meetings and its numbers are the full version of this page's Duvall half.

Somebody in this valley earns about $44,400 a year in a household of three, works at the school or the grocery or on a farm, and rents in Monroe because Duvall is out of reach. That household is the reason the state's number exists. It drives a commute the diesel price is quietly repricing and buys the groceries the index calls calm. It should hear the honest version from us: this zoning fight, as currently framed, will not produce a home it can afford. Upzoning changes what may be built. It does not fund anything, and the deepest band needs capital, not permission. What might actually reach that household is public land under the homes, the city's own fees waived, a rent ceiling that binds, subsidy the county and state control, and work close to home that pays enough to make the rent survivable. Those are separate decisions, made by different people, on different timelines, and the capacity showing will be filed as complete without any of them.

One household in this valley and its collaborators. We read the town's compliance artifact, tested it against the public record, and published what we found as a campaign on its own site, dontbothellduvall.com, with every source listed. The model behind its figures is published alongside it. We proposed putting the county's idle farmland back into production under a lease rather than a sale, with the jobs and food that follow. And we sell computing that a business, a farm, a nonprofit or a town hall owns outright, which is a thing we should disclose plainly: an argument that communities need their own means to check official arithmetic is an argument for something we sell. It does not make the argument wrong. It does mean you should check the numbers yourself, which is the entire point, and every source is listed below.

Two more honest notes. We are not against housing people, and this argument is borrowable by people who are. Our objection is to a regime that produces paper capacity for households it names while the public land that could house and employ those same households sits idle under designations nobody performs. And the farm proposal adds not one unit to the city's showing, because it lies outside the city limits. It addresses the real problem rather than the paper one, and we say so first. Any farm use of that ground turns on two things not yet settled: whether water and septic are feasible on the site, and whether the county determines the use compatible with the purpose it holds the land for.

• Duvall figures, in full with every citation: dontbothellduvall.com (the article and its 57 sources), the numbers and the record. The principal records follow.
• Washington State Department of Commerce, Petition for Review, Growth Management Hearings Board, Sept 3 2025. BERK Consulting, Land Capacity Analysis for City of Duvall Proposed Rezones, Dec 3 2025, Exhibits 4 to 6. City of Duvall, "Workshop on Duvall Municipal Code Title 14," slides, Sept 15 2026. Washington Office of Financial Management, April 1 housing estimates, Duvall. King County Countywide Planning Policies (Ordinance 19660).
• HUD FY2026 income limits, King County. City of Duvall draft DMC chapters 14.47 (affordable housing), 3.12 (multifamily tax exemption) and 14.20 (mixed use), Sept 1 2026. RCW 36.70A.540. Washington Joint Legislative Audit and Review Committee, Multifamily Tax Exemption review, 2019. Rent listings and our cost model, Sept 19 2026 (our estimates, marked as ours).
• City of Duvall Housing Action Plan, Sept 19 2023 (policy H-4.10, Actions I.1 and I.8); draft comprehensive plan, Sept 2026 (H 4.10, LU 7.1, LU 7.5, LU 9.2); City Administrator's memo "Fee Waiver Requests," Sept 8 2026; 2027-2028 preliminary budget presentation, Sept 15 2026.
• The Big Rock lot: Development Agreement, City of Duvall and CamWest, 2007-2008; statutory warranty deed, King County recording 20230111000142; King County Assessor, parcel 7299021190. City Council, Aug 19 2025, "Municipal Civic Campus Site Selection" (recording and minutes).
• The Urbanist, "Duvall Growth Plan Faces Challenge After Disregarding New Housing Law," Oct 17 2025 (the senior planner's letter to the state).
• City recordings: Planning Commission, Aug 26 2026 (automatic captions checked against the audio before use; the city published this recording under a September 1 Council title); City Council, Jan 6 2026 (about 3 hours 39 and 3 hours 50 minutes; the missing-middle explanation at about 4 hours 2 minutes); City Council, Sept 15 2026.
• Commuting: City of Duvall Housing Needs Assessment, 2023 (Census OnTheMap, 92.2 percent of working residents work outside the city). Flooding: Snoqualmie Valley Record, December 2025 road-closure reports. Parking: our sketch at standard rates.
• Growth Management Hearings Board, case 25-3-0003 (Futurewise and residents v. Mercer Island), Aug 1 2025; City of Mercer Island ordinances of July 21 2026.
• RCW 36.70A.302 (invalidity) and RCW 36.70A.340 (withheld revenues, suspended excise authority). Oregon HB 2001 (2019); Florida Live Local Act; California SB 79; American Planning Association 2026 survey on state limits to local accessory-dwelling authority. Leonard v. Bothell, 87 Wn.2d 847 (1976), on zoning and local referendum.
• King County: parcel 0626079039 and warranty deed 20201223002074; 2024 Regional Trail Needs Report (SNO-1, #31); Farmland Lease Program briefing to the Agriculture Commission, Feb 10 2022. The full record is on the farm proposal.
• Data centers: county and city moratoriums in Texas, Nebraska and Missouri, 2026; Georgia HB 1012; industry trackers on projects disrupted by local opposition, first quarter 2026.
• Public money: GAO, "Payment Integrity: Agencies' Estimated Improper Payments Increased to $186 Billion in Fiscal Year 2025" (64 programs, 15 agencies, up $24 billion year on year); GAO's estimate of $233 billion to $521 billion in annual federal losses to fraud, from fiscal 2018 to 2022 data; GAO on data quality and workforce for applying artificial intelligence to fraud and improper payments.
• Diesel: AAA national average, $6.23 per gallon on Sept 14 2026, an all-time high against the prior record of $5.78 in June 2022. Food at home up about 31.9 percent since January 2020 (BLS CPI food-at-home series through July 2026). QatarEnergy and other Gulf force majeure declarations on urea and ammonia, March 2026; industry estimates of about 40 percent of the urea export market disrupted; CRU, Argus and World Bank on a fertilizer-to-food transmission lag of roughly six to nine months; USDA WAOB and ERS on the 2026/27 wheat forecast.
• The frontier argument, the September 2026 pacing essay, the Ban Artificial Superintelligence Act and the July 2026 Hugging Face incident are sourced on the deeper page.
• Public opinion: Johns Hopkins (June 2026), Quinnipiac (March 2026), and the Carleton-CIGI-Ipsos survey on regulatory frameworks.

Working analysis by one valley household and its collaborators. Not legal, planning or financial advice, and not the position of any agency.