Critique: Nobody Wants Data Centers in Their Backyard

Source: jacobin
Authors: ByDavid Sirota
URL: https://jacobin.com/2026/05/ai-data-center-opposition-economics

What the article reports

A new Gallup poll showing majority opposition to data centers in local communities is the news peg. The piece argues the opposition is rational because data centers impose noise, raise utility costs, yield few permanent jobs, offer murky tax benefits, and power AI systems the public dislikes. It concludes local political backlash against data center deals is predictable and deserved.

Factual accuracy — Partial

The Wall Street Journal quote and the Synergy analyst figures (1,000+ construction jobs, fewer than 200 permanent) are specific and attributed: "rarely need more than one or two hundred once they open, according to Synergy chief analyst John Dinsdale." That is the piece's strongest evidentiary moment.

However, several causal claims are stated as fact without sourcing:

• "there is the perception — and some evidence — that data centers are linked to those price increases" — no study, jurisdiction, or figure is cited for the "evidence" half of this hedge.
• "a city council being voted out for advancing a data center deal" — the specific city is unnamed and no link is provided; unverifiable as written.
• The claim that data centers power "predatory pricing schemes" and are "ruining the education system" appears in a list of asserted harms with zero supporting evidence.
• The Gallup poll is described but no topline numbers, sample size, or margin of error are given, making the claim "most Americans hate the prospect" unreviewable.

Score pulled down by multiple unanchored causal assertions and the opaque Gallup citation.

Framing — Tendentious

1. "tech oligarchs creating these dystopias" — "oligarchs" and "dystopias" are connotation-heavy political epithets, not descriptive terms; using them in authorial voice signals advocacy rather than analysis.
2. "smug comment ignores" — the Twitter economist's argument is dismissed with an ad-hominem adjective ("smug") rather than engaged on its merits.
3. "all to enrich a distant tech bro's AI product, which is perceived to be dumb chatbots and clickslop videos" — "tech bro," "dumb chatbots," and "clickslop" are editorial judgments presented in the article's own concluding voice, not attributed to any source or poll respondent.
4. "entirely predictable for a few reasons worth reviewing, even if they are obvious" — the framing pre-empts any counterargument by characterizing the pro-data-center position as a failure to grasp the obvious.
5. "the perception is that it's all downside and very little upside" — the piece slides between "the perception" (poll-grounded) and its own stated view without marking the boundary; the rest of the paragraph treats the perception as established fact.

Source balance

Ratio: 2 critical/neutral : 0 genuine defenders of data center development. The one pro-data-center voice (the Twitter economist) is named only by platform and role ("prominent"), not by name, and is dispatched as "smug" rather than substantively addressed. No local official who favored a deal, no data center industry spokesperson making an affirmative case, no independent economist assessing net fiscal impact is quoted.

Omissions

1. Gallup topline numbers. The poll is the entire news peg; not reporting the percentage opposed, sample size, or question wording prevents any independent assessment of "most Americans hate."
2. Historical precedent for industrial opposition. The piece invokes the factory analogy but does not examine whether communities near data centers have outcomes meaningfully worse than comparable industrial sites — that comparison would test the core argument.
3. Counterevidence on tax revenue and property values. The piece acknowledges the tax-revenue argument exists but calls it "murky" without citing any study or jurisdiction where data centers have produced net positive fiscal outcomes — omitting the strongest version of the opposing case.
4. State and local incentive variation. Tax-break regimes vary enormously by state; treating them as a uniform negative generalizes from some cases to all.
5. Renewable-energy and grid-stability arguments. Some data center developers argue their facilities underwrite renewable buildout and grid investment. This goes entirely unmentioned.
6. The specific "city council voted out" case. No city, date, or outcome data is provided; a reader cannot verify or contextualize the claim.

What it does well

• The WSJ/Synergy job-figure citation — "rarely need more than one or two hundred once they open" — is the kind of specific, attributed data point that grounds an argument and is checkable.
• The nuclear-plant comparison ("nuclear power plants at least offer the possibility of more power produced with less particulate pollution") is a concrete analytical move that sharpens the distinction rather than relying purely on assertion.
• The piece is short and readable; its core claim (poor job-creation numbers undercut the standard economic-development pitch) is a legitimate journalistic observation, even if underdeveloped.

Rating

Overall: 4/10 — A piece with a defensible factual core that functions as advocacy rather than analysis, relying on loaded language, near-total source imbalance, and unattributed causal claims to carry arguments the evidence only partially supports.