Who benefits and who pays

The information architecture documented in the source article runs as follows. A developer secures a pre-application commitment from a local official under non-disclosure terms; the public learns of the arrangement through open-records requests; residents respond with the only remedy mapped onto the zoning-decision calendar — the recall petition. The sequence is reported across four jurisdictions: Lenox Township, Michigan (a rural municipality about 40 miles north of Detroit); Festus, Missouri; Yukon, Oklahoma; and Luther, Oklahoma.

In Lenox, the promotional site Lenoxdatacenter.com, which went live in May 2025, did not disclose the developer; emails obtained by residents through open-records requests showed developers had reached out to the township supervisor and deputy supervisor before any public application existed. The article applies the term “perceived secrecy” to the residents’ reaction. “The community still has questions that aren’t being answered and the public deserves to have transparency,” a resident said at a June board meeting, after trustees did not extend a four-month moratorium on data center development. Activists submitted a petition to recall four members of the Lenox board of trustees; public meetings during the period sometimes lasted more than four hours.

In Yukon, the city manager signed an NDA with Beltline Energy in May 2025; the council later voted four to one to sell the land to Beltline. The proposed project is valued at $1 billion. Vice Mayor Jeff Wootton, who resigned after a recall petition was filed, said the intent of the NDA “was never to conceal information from the public indefinitely but rather to protect sensitive negotiations during the preliminary stages,” and that he believed the data center could “bring significant long-term economic benefits to Yukon.” Joe Horn, “a Republican and bank vice president,” filed a separate petition against the Yukon mayor and vice mayor, citing concerns about “an alleged lack of transparency and water usage.” Horn stated: “We already ration water here in Yukon” and “at what cost are we going to have this huge industrial building right across the street from one of our most beautiful neighborhoods?”

In Festus, the city council approved a $6 billion agreement with developer CRG Clayco, including a five-year personal property tax abatement, after the developer had been rebuffed in St. Charles, where residents voted to ban data centers. The operating company was not named in the record on which the council voted. The mayor said the data center would bring “tremendous benefits to the people of Festus.” A judge determined that residents had gathered enough signatures to trigger a recall election against the mayor and three council members, but the city council rejected the petition; Dennis McDonald, “a Festus resident who teaches U.S. history at a local college,” filed a legal challenge. McDonald told the article: “If the community says: ‘Hey, we want to be involved and know what is going on here,’ and you just don’t involve them at every turn, it ends up building into where we are at now.”

“The company usually goes public only after the decisive votes have been taken,” Michael Bommarito, “an entrepreneur and author of ‘How to Fight a Data Center,’” characterized the developer’s timing directly. The information asymmetry, on this reading, functions as a first-mover advantage: it locks in local regulatory concessions — land purchases, council resolutions, tax abatements — before a coalition of opposition can form.

The strategic position of each party follows from this architecture. The developer holds a credible outside option: when one jurisdiction resists, the developer relocates. The article documents this directly — CRG Clayco moved from St. Charles to Festus; Beltline’s counterpart in Luther signed an NDA but, after observing Yukon’s reception, the Luther city council in June passed a six-month moratorium on data center development. The local official can trade near-term economic commitments for the political cost of eventual disclosure. The resident is place-bound and can respond only through open-records law and recall — what the article’s account frames as asymmetric political leverage that shifts the battleground from the project’s economic evaluation to the political survival of the officials who cast the decisive votes. In game-theoretic terms, the dynamic has been characterized as a local prisoner’s dilemma, in which individual jurisdictions may lower regulatory or tax standards to secure the projected economic base, fearing the loss of the investment to a neighboring municipality. The developer’s promises of “tremendous benefits” and “significant long-term economic benefits” function, in the same analytical frame, as what game theorists call “cheap talk”: the agreements as reported contain no commitment device tying the operator to specific jobs, water-use figures, or tax-revenue numbers over the abatement period, and the operating company is not named in the record on which the council voted. The resulting standoff reflects, in the article’s account, a broader anxiety over the physical footprint of artificial intelligence, in which the macroeconomic promises of the technology are mediated through the microeconomic and spatial realities of host communities.

What happens next

According to Data Center Watch, at least 75 data center projects worth about $130 billion were blocked or delayed in the first quarter of 2026 — a figure matching the full-year 2025 total. In May and June 2026, voters in California, Florida, Michigan, Missouri, Oklahoma, Oregon, and Texas launched recall efforts over data center handling, per the source article.

The recall threat has materialized in the article’s account: Wootton resigned after the Yukon petition was filed; the Festus council is in litigation over a judge’s signature ruling; activists in Lenox filed a recall petition against four trustees; and Data Center Watch’s 75-project count indicates the threat is being priced into site selection. The credibility of the resident move, in the terms the analysis requires, is high because it is no longer hypothetical.

The article reports that the United States has more than 4,400 data centers, per Data Center Map. A single facility can consume as much electricity as 2,000 homes, according to a University of Michigan report cited in the source. Typical data centers use about 300,000 gallons of water per day for cooling; large centers can use an estimated 5 million gallons daily — equivalent, according to the Environmental and Energy Study Institute, to the water use of a town of 10,000 to 50,000 residents. In Yukon, where water rationing is the baseline condition Horn describes, the question of what 300,000 or 5 million gallons means is, in the article’s own account, not abstract.

The pattern’s self-reinforcing character is documented directly: Luther’s moratorium followed Yukon’s backlash; the Festus agreement was reached only after St. Charles had refused; and the Lenox recall was filed during the same period in which the four-month moratorium lapsed. Each new case appears, on the substrate, to teach the next jurisdiction how to price the recall threat. The mechanism the article documents is the procedural correlate of public anxiety, in the words of Evan Sutton, “a Seattle resident who works in strategic communications and has volunteered to help data center opponents in 10 states”: “It reflects the growing anxiety about AI writ large” and “people feel like this technology is being shoved down our throats.”

How this is being framed

The source article applies the term “perceived secrecy” to the residents’ reaction to the information gap. Wootton’s counter-framing — that the NDA’s intent was to “protect sensitive negotiations during the preliminary stages” — illustrates the dispute over the duration of the information asymmetry rather than its existence. The strategic content of the NDA is, in this reading, the duration of the gap, since the public’s ability to organize opposition is bounded by what residents can learn through open-records requests and the calendar of decisive votes. The Festus and Yukon operators’ anonymity in the public record is itself a node of the same asymmetry the recall mechanism responds to: the operating company is the party whose identity residents cannot price, and whose actual footprint the negotiations have not been required to fix.

The opposition is bipartisan in the article’s account, which says the movement “has drawn together residents across party lines.” Anthony Leiserowitz, “director of the Yale Program on Climate Change Communication,” characterized the cross-partisan split in motivating concerns: “Many Democrats, especially those that are very environmentally minded, are deeply concerned about not just the energy use, but the use of polluting forms of electricity, like the re-invigoration of coal-fired power plants,” whereas “Republicans and, frankly, many Democrats as well, are concerned about the potential economic consequences, especially on their own electricity and energy bills.” A coalition organized around the latter concern is consistent with the article’s description of Horn as “a Republican and bank vice president” and of Festus, “a city in a county where Donald Trump won 67% of the vote in 2024.”

The spatial framing is also at issue. The article’s concrete descriptions — four-hour public meetings in a rural township roughly 40 miles north of Detroit; a small Missouri city whose council approved a $6 billion agreement with the operating company unnamed in the record; an Oklahoma town where water rationing is the baseline condition — ground the place-defending response in concrete geography. Christopher Alexander’s pattern language, applied at the settlement scale, treats the proposed placement as an extension of the intimacy gradient (Pattern 127) from interior building circulation to the neighborhood: a massive, resource-intensive industrial node directly adjacent to residential refuge. Jay Appleton’s habitat theory treats the resulting structure — typically a windowless, heavily secured monolith — as a collapse of the neighborhood’s prospect-refuge balance, replacing a human-scaled residential environment with an opaque industrial edge. Christian Norberg-Schulz’s phenomenological framework, as articulated in Genius Loci: Towards a Phenomenology of Architecture (1979), reads rural communities as deriving their character from a legible relationship between the man-made place and the natural place; on that reading, a data center operates as what the framework characterizes as a “placeless” extraction node, requiring only local resource inputs — electricity and water — without functional need for the local character, effectively overwriting the spatial character with an industrial energy sink.

The article documents a procedural pattern with enough specificity to ground the analysis. Its principal reporting gap, in the substrate itself, is the absence of project-specific commitments on operational water and energy load and of identification of the ultimate operating company in the three named cases. This gap is itself an analytical feature of the information asymmetry: the operating load is the variable on which residents’ electricity bills, water budgets, and quality-of-place claims depend, and it is the variable the negotiation has not been required to fix. Read in the article’s own sourcing, the analytical claim is that the recall wave is the downstream consequence of a site-selection strategy in which the developer’s outside option and the official’s electoral horizon are aligned against a resident constituency that learns from the precedent its neighbors have already set — a self-reinforcing cycle in which each new case appears, based on the reported record, to teach the next jurisdiction how to price the threat.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Balanced Critique
Weighs a proposal’s strengths and weaknesses evenhandedly.
Genius Loci — Sense of Place
Reads the character and felt quality of a place.
Strategic Interaction (Game Theory)
Models a situation as a game — players, moves, payoffs, and likely equilibria.
Loss Aversion
Losses loom larger than equivalent gains, skewing choices toward the status quo.