Texas turns public consent into a queue condition

About 90% of the 474 GW of proposed new load requests in Texas trace to data centers, a figure that reads as a growth story in a pipeline chart but as construction traffic, new noise, and open land turning into a job site from the kitchen window of a smaller community where one lands. Add the suspicion that the largest new consumer of power on the local grid is why the bill went up, and that is the view Gov. Greg Abbott handed the industry in August: he temporarily halted approvals for new data center projects inside Texas' grid interconnection process and ordered an audit of pending developments, citing grid reliability and the extraordinary volume of proposed new demand.

None of that should surprise anyone who has watched county boards in Virginia, but the venue changes it: Texas is one of the most pro-development states in the country and, in the commentary's own framing, expected to become one of the most important data center markets in the world, and it has now signaled that expansion will proceed on terms that protect the existing grid and the communities already relying on it. Craig Thompson, co-founder of Allegiance Search, an Austin-based executive recruitment firm whose practice covers digital infrastructure, energy and advanced manufacturing teams, makes the argument that this is the industry's next bottleneck in Construction Dive.

Recruiters read constraints through the roles they are asked to fill, which is why Thompson's framing deserves more weight than a typical op-ed: public support has become an input to development, sitting beside powered land and capital rather than trailing them as communications work. His evidence is the view from the fence line, where residents are weighing what changed outside their door — traffic, noise, land, energy costs — rather than artificial-intelligence compute demand or the long-run value of hyperscale infrastructure. If they believe the new load is pushing up their bills or straining infrastructure they already depend on, telling them the project will ultimately be good for the local economy does not win the argument; characterizing that as people failing to understand the opportunity, he writes, would be a mistake.

The Texas episode moves beyond a neighborhood fight because of where the objection now sits. Abbott's intervention carried it out of local planning meetings and into state infrastructure policy, where the questions are whether the proposed demand is realistic and whether the grid can serve it without placing an unreasonable burden on existing customers. Those are two different screening tools, and they favor different developers. The first rewards engineering and honest forecasting; the second is a ratepayer question, answered in public by whoever is willing to argue that the load pays for itself — a skill few developers have built in-house.

This publication's reporting has traced how quickly that screening has hardened: the halt arrived with a deadline attached, because ERCOT's December filing is meant to separate queued capacity from real demand, and 49.8 GW of Texas projects now wait on it. As argued here when the freeze landed, an audit turns a permit into a disclosure document and grid access into something closer to a state grant — an asset whose value now rests on a political judgment as much as an engineering one. That is a different instrument than the interconnection position a developer thought it held last spring, and in Texas the paperwork no longer speaks for itself.

The house view on this beat has been that consent is the product in every data center trade, and Texas sharpens the point: public support is not a soft constraint sitting outside the capital stack; it is a schedule input, and schedule is what decides whether an anchor-contracted data center earns infrastructure pricing or drifts toward a merchant bet. Merchant risk on a queue position is worse than merchant risk on a power contract, because the state that granted the position can suspend it. The sector's capital hierarchy already sorts digital infrastructure by whether revenue is contracted and whether the tenant is a hyperscaler; Texas has added a second gate, whether the state and the neighbors will let the electrons move at all. For projects large enough to threaten the local rate base, consent prices before capacity.

None of this argues that power stops mattering; grid capacity, interconnection queues and load-class rules remain the binding constraint on the buildout from data centers to factories, and nothing in the Texas episode changes that. The change is in the number of locks on the door: supply-side policy can add megawatts, but it can also revoke a schedule, and Texas has just shown it will. For developers, community and ratepayer positions now belong in the same diligence binder as the interconnection agreement, because in a queue that has become partly political, projects with negotiated local support should clear an audit faster than projects with nothing but the best land. That is a testable claim, and Texas is about to test it.

December's filing is where the argument meets an auditor. It will establish how much of the 49.8 GW still qualifies as real demand under the state's scrutiny, and for the developers holding those positions, whether the past year went into building consent or assuming it. The number worth watching is how many of the surviving projects can point to a community that agreed to host them, not how much capacity survives the audit.