Blog

How Many Jobs Does a Data Center Actually Create?

jobs economic development JLARC tax breaks subsidies Virginia construction clawbacks community advocacy explainer

Every data center pitch leads with jobs — usually thousands of them. The number that actually matters is far smaller and fully knowable: a typical hyperscale hall runs on about 50 full-time workers, half of them contractors. Here's the real math from Virginia's own legislative watchdog, the construction mirage behind the big headline, and the three questions that put it on the record.

$ PUBLIC COST

Every data center proposal leads with jobs. The banner number is almost always in the thousands. The number that matters for your community is far smaller — and, unlike most of the pitch, it is precisely knowable.

The claim versus the reality

Virginia's Joint Legislative Audit and Review Commission (JLARC) produced the most thorough independent study of the industry to date. Its finding on permanent employment:

A typical 250,000-square-foot data center employs about 50 full-time workers — roughly one job per 5,000 square feet, and about half of those are contractors, not the operator's own staff.

(JLARC, Data Centers in Virginia, 2024.)

Fifty people. For a building the size of several football fields drawing the power of a small city. That is the operating reality of a facility designed, by definition, to run on automation.

And those fifty aren't the office jobs the pitch implies. A running data center employs security staff, HVAC and mechanical technicians, a few electricians, and a small network/facilities crew. The servers are owned and operated remotely by the tenant; the "high-tech jobs" mostly live at a corporate campus in another state. What lands in your county is a hardened, lightly-staffed industrial building.

The construction mirage

So where do "thousands of jobs" come from? Construction. A large build can peak at as many as 1,500 workers on site — but the build lasts just 12 to 18 months, and then those workers leave for the next site. A headcount that disappears in eighteen months is not a tax base, and it is not what "jobs" means to a family deciding whether to support a rezoning.

The statewide illusion

JLARC credits data centers with 74,000 jobs, $5.5 billion in labor income, and $9.1 billion in annual GDP across all of Virginia. Those numbers are real — but the majority is construction and indirect activity, not permanent operations. When a developer quotes a statewide industry total to describe your one facility, that is the sleight of hand to watch for. The state's whole data center sector is not moving into your county; roughly fifty operating jobs are.

The number nobody puts on the slide: cost per job

Here is where it stings, and it's not a hypothetical — the per-job math has been done. Good Jobs First found that eleven of the largest data centers received public subsidies worth about $1.95 million per permanent job. In Ohio, the per-job cost runs roughly $1.4 million at Google's facilities and $1 million at Meta's. For comparison, Good Jobs First recommends states cap data-center subsidies at $50,000 per job — the deals on the table today routinely run twenty to forty times that.

Two facts make it worse. First, the giveaway is often unnecessary: Georgia's own analysis concluded that about 70% of data-center projects would have located there even without the subsidy — the state is paying billions for investment it was already getting. Second, the largest state exemptions now cost over $1 billion a year each, and Georgia alone raised its projected FY2026 loss to $2.5 billion. The jobs are real. They are just, per job, among the most expensive economic development a state can buy.

The move: three questions for the record

Don't argue the banner number — pin down the real one. Ask these at the hearing, and ask for the answers in writing (the hearing-prep page has more):

  1. How many permanent, full-time jobs employed by the operator — not contractors — and will that floor be guaranteed in the agreement?
  2. What is the total public subsidy, and the subsidy per permanent job?
  3. Is there a clawback if the promised job numbers aren't met?

Then set the answer next to the tax breaks on the table and run your own local figures on the impact calculator. A project that genuinely creates lasting jobs will answer all three without flinching. Most can't.

← PreviousHow to Unmask the Shell Company Behind a Data CenterNext →"The A.I. Revolt Is Here": When Ezra Klein Covers Your Zoning Fight, the Politics Have Changed