The Excavation Industry Comes for the Data Center
Who profits when the server farm never gets built
Drive past a proposed data center site in almost any state right now and you'll find the same scene. A county commission meeting running four hours past its scheduled end. A line of residents at the microphone raising a real concern — the water table, the noise, the tax abatement, the power bill. A team of consultants and attorneys in the back row, billing by the hour regardless of which way the vote goes.
The concerns are almost always legitimate. The water really can drop. The transformer hum really does carry at night. The tax abatement really does shift the burden onto the school district. None of that is manufactured.
What's worth asking is a different question: who benefits from the fight lasting as long as possible, regardless of who wins it?
The Objections, Inventoried
Strip the data center fight down to its component objections and you get a short, consistent list, repeated from Loudoun County to rural Georgia to the Ohio corridor:
Grid strain — new industrial load competing with residential ratepayers for capacity
Water usage — cooling systems drawing down aquifers, especially in already water-stressed regions
Noise — round-the-clock cooling fans and backup generators near residential zones
Land use — farmland or green space converted to industrial footprint
Tax abatements — locals subsidizing a facility that pays little in ongoing property tax
Low job yield — a construction boom followed by a handful of permanent, often non-union technical jobs
Diesel backup generators — air quality concerns from emergency power systems
Loss of local control — decisions made by state economic development offices over local objection
Every one of these is defensible on its own terms. But run each through the question the original excavation industry argument insists on: not "is the concern real," but "who has a financial position staked on how this plays out, and did they place that position before the fight started."
Money, Traced
The answer is different for almost every line item — which is itself the finding.
Grid strain cuts both ways. Utilities often want the load, since new interconnection triggers rate-base expansion they earn a guaranteed regulatory return on. Opposition tends to come from incumbent generation owners who don't want new capacity competing with theirs, or utilities angling to negotiate premium contract rates before agreeing to connect.
Water usage traces to agricultural and municipal water-rights holders with existing allocation contracts, for whom a new industrial draw threatens a negotiated share they already depend on.
Land use and zoning trace to competing jurisdictions — counties and states actively bidding for the same project with rival incentive packages, sometimes amplifying local opposition in a jurisdiction they're trying to out-recruit.
Tax abatements trace most cleanly of anything on the list — school districts and municipal governments losing real property tax base, a direct and quantifiable loss, arguably the most legitimate financial objection here.
Diesel generators and environmental review trace to the same actor the original piece named in the California high-speed rail case: environmental attorneys, expert witnesses, and permitting consultants who are paid by the hour, for whom the length of the fight — not its outcome — is the revenue.
That last category is the closest thing to a single node connecting the whole list. Not one industry with one motive, but a permitting-and-review complex with a shared interest in prolongation, regardless of which side eventually wins.
The Precedent Already Exists
This isn't a new pattern invented for data centers. It's the same mechanism that already reshaped American energy policy once, and the paper trail is public.
In 1970, Robert O. Anderson — head of Atlantic Richfield, a major oil company — personally funded the founding of Friends of the Earth with a $200,000 donation. The organization became one of the most influential anti-nuclear voices in the country for the next fifty years. Coal industry leadership was on record even earlier, in 1963, warning that continued federal investment in nuclear power risked destroying the coal industry outright — not a hidden motive, just a competitor saying the quiet part in an internal memo. In the decades since, investigative reporting has surfaced branded fossil-fuel campaign material designed to manufacture the appearance of grassroots anti-nuclear sentiment where the underlying sentiment wasn't organic.
None of this means the public's fear of nuclear power was fake. Three Mile Island and Chernobyl were real events with real consequences, and the fear that followed was genuine. But genuine fear and funded amplification aren't mutually exclusive — the fossil fuel industry didn't invent the fear, it invested in keeping it alive long after the safety data stopped supporting the scale of it. The capacity gap nuclear left behind got filled almost entirely by natural gas, not renewables, for the next three decades.
Coal miners' unions were part of that same 1960s coalition opposing nuclear, and their motive requires no conspiracy theory at all: a new technology threatened their members' jobs, so their union did exactly what unions exist to do. The same logic ran in reverse when nuclear plants themselves faced closure — nuclear plant workers in Zion, Illinois made the identical argument coal workers had made a generation earlier: high-wage union jobs, an economic engine for the town, a tax base the community depended on. Different fuel, same defense, same institutional reflex. Whoever's labor is already attached to the incumbent system defends it — not out of ideology, but because that's the job description.
The One Actor With Both Keys
Here's where the data center fight differs from every other example in this piece.
Land use, water permits, and tax abatements get decided locally — county commissions, planning boards, city councils. That's where the permitting-and-review complex operates, and it's where most of the "excavation industry" friction against data centers currently lives.
But unions are the one actor in this entire chain with genuine standing at both levels simultaneously. A local IBEW hall can turn out members and testify at a county zoning hearing the same way coal miners' locals did in the 1960s. The national union can apply pressure through state legislation, NLRB posture, and federal permitting policy at the same time. No permitting consultant has that reach. No competing jurisdiction has that reach. Utilities are regional at best. Unions are the only actor positioned to move both the county commission and the statehouse from the same institutional base.
Which raises the actual lever available here.
Transmission line construction and power plant construction are already heavily unionized — IBEW on the lines, building trades on the plants. If that's true, and if grid interconnection and generation buildout are the real binding constraints on data center growth (which most reporting on the industry suggests they are), then the labor building the actual bottleneck infrastructure already has a direct financial stake in projects moving forward, not stalling.
What's missing is the operational workforce — the permanent staff running the finished facility, currently thin and largely non-union. Right now that's also the industry's weakest political argument: few permanent jobs, easy target for the "low job yield" objection. Organize that workforce, and the calculus changes on both ends. The unions with clout over the harder bottleneck — transmission and generation — already have skin in the game. Add an organized operational workforce with its own stake in the facility's survival, and organized labor stops being a neutral bystander to data center fights and becomes an active political counterweight to the permitting complex profiting from delay.
The TRIZ Forward Read
The underlying contradiction here is the same one that runs through every example in this piece: the people with the technical and legal standing to resolve a dispute quickly are rarely the people whose revenue depends on it staying unresolved. Engineers sign PE stamps and carry personal liability. Consultants bill hourly with no such exposure. That asymmetry doesn't fix itself.
But unions occupy a different position in that asymmetry than any other actor in the data center fight. Their institutional survival depends on organizing new industrial workforces, not on prolonging permitting fights — the opposite incentive from the consulting-and-review complex. If that incentive gets pointed at data centers specifically, the likely trajectory over the next several years is a gradual shift from adversarial-or-neutral labor relations toward the same alliance nuclear plants and large industrial builds already have with their unionized workforces: an organized, well-funded political constituency arguing for speed, not against it.
The lawn is still there. The excavators are still bidding. But this time, one of the trades on site has a reason to want the job finished — not just started.
Herbert Roberts, P.E. spent 32 years in aviation R&D across two companies and has spent the last eight years analyzing accidents for attorneys under his PE license, translating engineering findings into legal language.
Herbert Roberts, P.E. — 30+ years in aviation R&D across two companies. 8+ years analyzing accidents for attorneys under my P.E. license.