A false scientific claim, misused noise data, activist rhetoric and no Berkeley Heights-specific analysis are no basis for a townwide ban.

AI data centers deserve serious scrutiny. A large proposed facility should be examined for electricity demand, water use, noise, emissions, infrastructure costs, tax revenue, jobs and proximity to homes.

But Susan Poage's presentation does not provide that analysis. Instead, it moves from unspecified "rumblings" about the Nokia property to support for a Township-wide ban. The deck identifies no applicant, project size, electrical load, cooling system, water estimate, generator plan, sound study or fiscal analysis.

That is not responsible research. It is a conclusion in search of evidence.

Poage has not demonstrated the expertise required for this report

Poage's official Township biography describes her as a longtime kindergarten and first-grade teacher, a recipient of a Presidential Award for Excellence in Mathematics and Science Teaching, a grant writer and a member of the Planning Board. Those are legitimate accomplishments.

But the biography identifies no professional experience in data-center engineering, electrical-grid planning, industrial cooling, hydrology, acoustics, air-quality modeling, utility economics or large-scale commercial development.

Being a teacher does not prevent someone from studying a technical subject or forming a policy opinion. But when an elected official lacks specialized expertise, the answer is to retain qualified professionals, disclose every source and carefully distinguish established findings from speculation.

Poage did none of that here. She presented an amateur slideshow with the confidence of a technical report without doing the work necessary to earn that confidence.

The "peer-reviewed research" claim falls apart

Poage says she researched "peer reviewed articles." Yet the 14-slide presentation has no bibliography, no reference page and virtually no slide-level citations. Its content can be traced primarily to government reports, policy explainers, a magazine article, journalism, a short journal letter and an interview with anti-data-center activists.

Government reports can be valuable. Policy articles can raise legitimate issues. But they are not all peer-reviewed scientific studies, and an activist interview certainly is not one.

Calling this source collection "peer reviewed research" gives the public a misleading impression of its rigor.

Example one: the water-cycle claim is scientifically false

The presentation states:

"Cooling water is not part of the water cycle."

That is simply wrong.

The U.S. Geological Survey explains that evaporation moves water from Earth's surface into the atmosphere and is a fundamental process in the water cycle. The Department of Energy similarly explains that cooling towers reject heat through evaporation.

There is a real issue here: evaporative cooling can represent consumptive local water use because the evaporated water is not promptly returned to the same utility or watershed. Cooling-tower blowdown can also require treatment and management.

Poage could have made that accurate argument. Instead, she chose a scientifically false slogan.

That is lazy. Worse, it causes informed residents to question every other technical statement in the presentation.

Example two: she misuses an internal noise measurement

Poage's health slide cites internal noise levels as high as 96 dBA, compares them with an occupational hearing threshold and immediately claims harm to nearby communities.

That is an apples-to-oranges comparison.

A noise measurement taken inside a server facility does not establish the sound level at a residential property line. Community exposure depends on the equipment, building enclosure, setbacks, barriers, terrain, operating schedule, low-frequency characteristics and distance from homes.

The official JLARC study that Poage links reaches a much more careful conclusion. It says low-frequency data-center noise has affected some nearby residents' well-being, but is generally not loud enough to damage their hearing. It also found that a large majority of facilities do not generate complaints because of their design or location. JLARC recommends sound modeling and enforceable zoning standards - not pretending an internal workplace measurement is a neighborhood exposure measurement.

Noise can be a legitimate reason to reject a poorly designed or badly located project. Poage's misuse of the 96 dBA figure is not a legitimate way to demonstrate that risk.

Example three: she conflates a server room with a hyperscale AI campus

The presentation portrays data centers as though they are all city-sized industrial complexes and then declares that even small facilities are a burden.

The Congressional Research Service says data centers range from closets and rooms inside ordinary buildings to massive hyperscale campuses. A hyperscale facility can exceed 100 megawatts - roughly the electricity demand of 80,000 households - but that plainly does not describe every facility classified as a data center. CRS also states that smaller facilities within office buildings might increase direct water use only incrementally.

JLARC likewise found that most data centers it examined used approximately the same amount of water or less than an average large office building. Some used substantially more; some used less than a typical household. The result depended on facility size, computing density and cooling technology.

That variation is exactly why a competent ordinance would define facility types, sizes and power thresholds. Poage's presentation erases those distinctions because acknowledging them would weaken the argument for a blanket ban.

Example four: the health claims outrun the evidence

Poage's presentation asserts that pollution associated with data centers is increasing respiratory disease, cardiovascular illness and cancer risk in nearby communities.

Pollutants from diesel engines and fossil-fueled generation are real health hazards. But establishing that a particular facility has increased disease rates in a neighboring population requires actual exposure data, generator operating hours, emissions controls, dispersion modeling, background pollution measurements and epidemiological evidence.

The wording in Poage's slide closely matches a 2025 publication titled Global Data Center Expansion and Human Health: A Call for Empirical Research. That publication is classified by the journal as a Letter, not an original epidemiological study. It contains only four references, including a technology website and an unpublished arXiv modeling paper.

Most damaging to Poage's presentation, the authors themselves say that little is known about long-term health outcomes in data-center host communities and that empirical research is urgently needed.

Poage took a source calling for more research and converted it into a slide suggesting the causal health effects were already established.

That is not responsible public-health communication. It is fear-based advocacy.

Example five: activist rhetoric is presented as "The Harsh Reality"

One slide declares that the beneficiaries of data centers are Big Tech and the surveillance, war and fossil-fuel industries. It labels this "The Harsh Reality," as though it were the conclusion of an objective economic or engineering analysis.

It was not.

That statement came from a Truthout interview with Lee Ziesche, an organizer with the Tucson Democratic Socialists of America and the No Desert Data Center Coalition. Ziesche was discussing opposition to a particular development in Tucson, Arizona.

Ziesche has every right to express that political opinion. Poage has every right to agree with it.

But an activist's statement about Tucson is not a factual finding about the Nokia property, Berkeley Heights' utility system or the fiscal consequences of a hypothetical local project. Removing the speaker, affiliation and project-specific context and then displaying the statement as an objective "reality" is exactly how propaganda works.

Example six: she cherry-picks the government report she actually links

JLARC's Virginia study is the strongest source associated with the presentation. It confirms serious concerns about electricity demand, residential siting, infrastructure costs and noise.

But Poage presents only the negative side.

The same report found that data centers:

JLARC recommends stronger zoning, sound modeling, water estimates, utility-rate protections and better cost allocation. It does not conclude that every locality should prohibit every type of data center.

Poage uses JLARC's warnings while concealing most of its balancing evidence and practical recommendations. That does not make every sentence in her presentation false. It makes the overall presentation profoundly misleading.

There is no Berkeley Heights analysis

The most glaring failure is the absence of local evidence.

Poage begins with "rumblings" of possible interest at Nokia. Her presentation does not identify an application, buyer, developer or proposed project.

The Township's official Nokia page says the land sale is private, the municipality is exploring a broad range of possible uses and zoning and redevelopment procedures give the Township leverage over future development. It does not identify a proposed data center or provide any facility specifications.

Before recommending a permanent prohibition, Poage should have answered basic questions:

What type of data center is supposedly being considered? How many megawatts? What cooling system? How much water? How many generators? How close to homes? What infrastructure upgrades? What tax revenue? What operating jobs? What sound level at the property line?

Her presentation answers none of them.

A blanket ban is not the only possible policy

Even the World Resources Institute article used by the presentation says community harms are not inevitable. They depend on decisions by developers and local governments. WRI discusses safeguards including water-use monitoring, drought plans, reclaimed water, financial protections, stronger siting rules and project-specific review.

Berkeley Heights could establish:

CRS also explains that electricity-rate effects are not automatic: new infrastructure can increase rates, while additional sales using existing capacity can sometimes spread fixed costs and reduce them. The outcome depends on the specific utility system and rate structure.

Poage evaluates none of these distinctions or alternatives. She jumps directly to prohibition because the presentation was created to sell a ban, not to determine the best policy.

The bottom line

Calling this presentation lazy is justified. It contains no bibliography, no local engineering work, no fiscal analysis, no meaningful differentiation among facility types and no serious consideration of regulatory alternatives.

Calling it propaganda is also justified. It selects frightening facts, suppresses balancing findings, misapplies technical evidence, presents activist opinion as objective truth and marches residents toward a conclusion that was apparently chosen in advance.

Susan Poage is entitled to oppose data centers. She is not entitled to have a technically careless advocacy deck accepted as competent research merely because she holds public office.

AI data-center proposals should be taken seriously. Any actual proposal that cannot satisfy strict environmental, infrastructure and financial standards should be rejected.

But banning an entire category of land use based on Poage's amateur presentation is nonsensical. Berkeley Heights deserves qualified analysis, transparent evidence and sober policymaking. On all three counts, Susan Poage failed the town.

Primary references