Opinion: A $50 Billion Data Center Is Not a Jobs Program

Meta's data center in rural Richland Parish, Louisiana, will eventually cost roughly $50 billion to build — more than triple the figure first floated when the project was announced. It will span about five square miles, draw five gigawatts of electricity once complete, and, when fully operational sometime around 2030, support more than 1,000 jobs.

Run the arithmetic and that works out to something in the neighborhood of $50 million of capital investment for every permanent job created. Local officials have welcomed the project warmly, and it's easy to see why in the short term: Meta has committed $1 billion to local infrastructure, including roads and water systems, and new tax revenue has already funded increased teacher bonuses in the parish. Those are real, tangible benefits to a rural community that has not always attracted this scale of private investment.

But it is worth being honest about what kind of project this actually is, because the framing matters for how other communities across the country evaluate the wave of similar mega-projects now being pitched to them. This is not principally a jobs program, however it gets described in press releases and ribbon-cutting speeches. It is closer to industrial infrastructure — more akin to a power plant or a refinery than a factory, in that its economic value comes overwhelmingly from the compute capacity and electricity it delivers to Meta's global operations, not from the relatively modest headcount required to keep tens of thousands of servers running.

That distinction should shape the terms communities negotiate. A project whose primary local benefit is tax revenue and one-time infrastructure investment, rather than durable employment, calls for a different kind of deal than a factory that will employ thousands of people for decades. It argues for locking in commitments on energy costs and grid capacity — since a five-gigawatt draw is a serious claim on regional electricity supply that could affect rates and reliability for ordinary residents for years, an issue already surfacing in several U.S. regions hosting large data center buildouts. It argues for firm, contractually binding infrastructure commitments rather than announcements that can quietly shrink. And it argues against selling tax incentives to constituents primarily on the promise of jobs that the underlying economics of the project were never going to deliver at meaningful scale.

None of this means these projects are bad deals, or that Louisiana was wrong to pursue one. Data center investment is going to keep flowing somewhere as AI infrastructure spending climbs into the hundreds of billions of dollars nationally, and a rural parish capturing a share of that, along with genuine infrastructure upgrades, is not nothing. But the public conversation around these projects — in Louisiana and in the dozen or so other states now competing for the next hyperscale campus — would benefit from more precision about what they are and are not. They are enormous industrial infrastructure investments with modest direct employment. Communities that understand that distinction going in are better positioned to negotiate for the things these projects can actually deliver — infrastructure, tax revenue, and protected electricity costs — rather than overselling the jobs number to a public that will notice, eventually, how few people actually work behind the fence.